Creating virtual patients using large language models: scalable, global, and low cost.

Virtual patients (VPs) have long been used to teach and assess clinical reasoning. VPs can be programmed to simulate authentic patient-clinician interactions and to reflect a variety of contextual permutations. However, their use has historically been limited by the high cost and logistical challenges of large-scale implementation. We describe a novel globally-accessible approach to develop low-cost VPs at scale using artificial intelligence (AI) large language models (LLMs). We leveraged OpenAI Generative Pretrained Transformer (GPT) to create and implement two interactive VPs, and created permutations that differed in contextual features. We used systematic prompt engineering to refine a prompt instructing ChatGPT to emulate the patient for a given case scenario, and then provide feedback on clinician performance. We implemented the prompts using GPT-3.5-turbo and GPT-4.0, and created a simple text-only interface using the OpenAI API. GPT-4.0 was far superior. We also conducted limited testing using another LLM (Anthropic Claude), with promising results. We provide the final prompt, case scenarios, and Python code. LLM-VPs represent a 'disruptive innovation' - an innovation that is unmistakably inferior to existing products but substantially more accessible (due to low cost, global reach, or ease of implementation) and thereby able to reach a previously underserved market. LLM-VPs will lay the foundation for global democratization via low-cost-low-risk scalable development of educational and clinical simulations. These powerful tools could revolutionize the teaching, assessment, and research of management reasoning, shared decision-making, and AI evaluation (e.g. 'software as a medical device' evaluations).

Effect of two esthetic digitally produced materials used in fabrication of extracoronal attachments on the stresses Induced in removable partial dentures.

BACKGROUND: Preservation of the remaining structures while maintaining an esthetic appearance is a major objective in removable partial prosthodontics. So, the aim of the current study was to compare the stresses induced on the supporting structures by two digitally produced esthetic core materials; Zirconia and Polyetheretherketone when used as an extracoronal attachment in distal extension removable partial dentures using strain gauge analysis. METHODS: A mandibular Kennedy class II stone cast with the necessary abutments' preparations was scanned. The mandibular left canine and first premolar teeth were virtually removed. An acrylic mandibular left canine and first premolar teeth were prepared with heavy chamfer finish line and scanned. Virtual superimposition of the acrylic teeth in their corresponding positions was done. Two strain gauge slots were designed: distal to the terminal abutment and in the residual ridge. Two models and two sets of scanned teeth were digitally printed. The printed teeth were then placed in their corresponding sockets in each model and scanned. The attachment design was selected from the software library and milled out of Zirconia in the model ZR and Polyetheretherketone in the model PE. Five removable partial dentures were constructed for each model. The strain gauges were installed in their grooves. A Universal testing machine was used for unilateral load application of 100 N (N). For each removable partial denture, five measurements were made. The data followed normal distribution and were statistically analyzed by using unpaired t test. P value < 0.05 was considered to be statistically significant. RESULTS: During unilateral loading unpaired t test showed statistically significant difference (p = 0.0001) in the microstrain values recorded distal to the abutment between the models ZR (-1001.6 µÎµ ± 24.56) and PE (-682.6 µÎµ ± 22.18). However, non statistically significant difference (p = 0.3122) was observed in the residual ridge between them; ZR (16.2 µÎµ ± 4.53) and PE (15 µÎµ ± 3.74). CONCLUSIONS: In removable partial dentures, Polyetheretherketone extracoronal attachment induces less stress on the supporting abutments compared to the zirconia one with no difference in the stresses induced by them on the residual ridge.

Efficacy of the nasal airflow-inducing maneuver in the olfactory rehabilitation of laryngectomy patients: A systematic review and meta-analysis.

INTRODUCTION: This is the first systematic review and meta-analysis to investigate the effectiveness of the nasal airflow-inducing maneuver (NAIM) in olfactory rehabilitation for total laryngectomy (TL) patients. METHODS: We conducted a systematic literature search following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The inclusion criteria required that patients must have undergone a TL with subsequent NAIM training for at least 2 weeks and olfactory evaluation. The impact of NAIM on olfactory outcomes compared to that at baseline was measured. Olfactory measures included the Sniffin' Sticks Test, Smell Disk Test, Scandinavian Odor Identification Test, and Quick Odor Detection Test. The primary outcome measures were the proportion of patients with normosmia at baseline and after intervention. RESULTS: Seven studies from 2000 to 2023 comprising a total of 290 TL patients met the inclusion criteria. The meta-analysis revealed that prior to intervention, the pooled proportion of patients with normosmia was 0.16 (95% confidence interval [CI]: 0.09‒0.27, p = 0.01). After intervention, the same proportion increased to 0.55 (95% CI: 0.45‒0.68, p = 0.001). Among the included patients, 88.3% were initially anosmic or hyposmic, which was reduced to 48.9% after NAIM practice, with 51.1% achieving normosmia. The percent improvement was not found to be significantly associated with the timing of intervention post-TL (p = 0.18). CONCLUSIONS: NAIM increased the proportion of patients who achieved normosmia in TL patients. NAIM stands out as a safe, easily teachable maneuver with promising results. Further efforts are warranted to provide specific recommendations and guidelines for the use of NAIM in clinical practice.

Characterizing the consistency of motion of spermatozoa through nanoscale motion tracing.

OBJECTIVE: To demonstrate nanoscale motion tracing of spermatozoa and present analysis of the motion traces to characterize the consistency of motion of spermatozoa as a complement to progressive motility analysis. DESIGN: Anonymized sperm samples videographed under quantitative phase microscope, followed by generating and analyzing super-resolution motion traces of individual spermatozoa. SUBJECTS: Centrifuged human sperm samples. MAIN OUTCOME MEASURES: Precision of motion trace of individual sperms, presence of helical pattern in the motion trace, and mean and standard deviations of helical periods and radii of sperm motion traces, speed of progression. RESULTS: Spatially sensitive quantitative phase imaging with a super-resolution computational technique MUltiple SIgnal Classification ALgorithm (MUSICAL) allowed achieving motion precision of 340 nm using 10×, 0.25 NA lens whereas the diffraction limited resolution at this setting was 1320 nm. The motion traces thus derived facilitated new kinematic features of sperm, namely the statistics of helix period and radii per sperm. Through the analysis, 47 sperms with speed >25 µm/sec randomly selected from the same healthy donor's semen sample, it is seen that the kinematic features did not correlate with the speed of the sperms. Also, it is noted that spermatozoa may experience changes in the periodicity and radius of the helical path over time. Further, some very fast sperms (for example >70 µm/sec) may demonstrate irregular motion, needs further investigation. Presented computational analysis can be used directly for sperm samples from both fertility patients with normal and abnormal sperm cell conditions. We note that MUSICAL is an image analysis technique which may vaguely fall under machine learning category, but the conventional metrics for reporting found in EQUATOR do not apply. Alternative suitable metrics are reported, and bias is avoided through random selection of regions for analysis. Detailed methods are included for reproducibility. CONCLUSION: Kinematic features derived from nanoscale motion traces of spermatozoa contain information complementary to the speed of the sperms, allowing further distinction among the progressively motile sperms. Some highly progressive spermatozoa may have irregular motion pattern, and whether irregularity of motion indicate poor quality regarding artificial insemination needs further investigation. Presented technique can be generalized for sperm analysis for a variety of fertility conditions.

Autotransplantation with Adjunctive Application of Enamel Matrix Derivatives Using a Digital Workflow: A Prospective Case Series.

OBJECTIVES: Digital protocols and bioactive materials may reduce complications and improve tooth autotransplantation (ATT) success and survival rates. This prospective study assesses the performance of a fully digital autotransplantation protocol of close-apex molars with the adjunctive application of Enamel Matrix Derivatives (EMD). METHODS: Twelve adult patients with 13 hopeless molar teeth were replaced with autotransplantation of closed apex third molars. Outcomes, including success and survival rates, clinical, endodontic, radiographic, patient-reported outcome measures (PROMs), and digital image assessments, were conducted over a two-year follow-up period. RESULTS: Survival and success rates were 100% and 91.2%, respectively, with no progressive inflammatory or replacement root resorption (ankylosis) except for one tooth presenting radiographic furcation involvement. A significant probing depth reduction of 2.4 ± 2.58 mm and CAL gains of 2.8 ± 3.03 mm were observed in transplanted teeth compared to the hopeless receptor teeth. Radiographic bone levels remained stable throughout the study period (-0.37 ± 0.66 mm), and digital image assessments showed minimal alveolar ridge width changes (-0.32 to -0.7 mm) and gingival margin changes (-0.95 to -1.27 mm) from baseline to last visit. PROMs indicated very high patient satisfaction. CONCLUSION: The use of a digital ATT protocol with adjunctive use of EMD in closed-apex third molars demonstrated promising short-term high success and survival rates. Additionally, this type of therapy adequately preserves the dimensions of the alveolar ridge in the receptor site.

Intraoperative Fluoroscopy Versus Navigation to Determine Cup Anteversion in Direct Anterior Total Hip Replacement: A Technical Trick for Obtaining "True" Anteversion.

During acetabular cup positioning, intraoperative measurements of cup anteversion were taken using both fluoroscopy and navigation system. With the C-arm introduced at 40°, an anteroposterior view of the pelvis is taken. The C-arm is then centered over the hip, showing an anteverted cup with an approximate inclination of 40°. The axial C-arm is tilted away until the cup opening is visualized as a straight line, indicating that the beam of the fluoroscopy is aligned with the cup's anteversion. The tilt angle on the C-arm and anteversion reading on the navigation workstation were recorded. The high degree of agreement between fluoroscopic and navigation measurement of acetabular cup anteversion supports the use of fluoroscopy in settings with limited access to navigation systems in direct anterior total hip arthroplasty.

One-stage prosthodontically driven jaw reconstruction in patients with benign and malignant pathologies: A 7- to 11-year cohort study.

OBJECTIVES: One stage functional jaw reconstruction is defined as the resection and reconstruction of segmental defects in conjunction with the placement of dental implants in an ideal prosthetic position and loaded with a provisional restoration, during one surgical procedure. The aim of the study is to describe clinical outcomes of patients who underwent one stage functional jaw reconstruction. METHODS: Patients who underwent one-stage functional jaw reconstruction, from January 2013 to March 2016 were recalled in 2022 and 2023. Planning and execution for the reconstruction utilized either analogue or digital techniques. Outcome parameters recorded were treatment-related outcomes at patient level, implant-related outcomes and patient-reported outcome measures. RESULTS: Eighteen patients underwent one-stage jaw reconstruction with a total of 57 implants. Four patients had maxillary and 14 had mandibular reconstructions. Ten patients underwent postoperative radiotherapy. Ten patients were planned using analogue and eight by digital planning. Three patients had partial flap necrosis, three patients had plate fractures, implant loss was seen in one patient and four patients died during the period. A functional prosthesis was provided in 16 out of the 18 patients. CONCLUSION: One-stage functional jaw reconstruction is a predictable method for providing rehabilitation with successful outcomes at 7-11 years. However, caution should be exercised when the treatment modality is carried out in patients with malignant pathologies who have undergone radiotherapy.

NextLens-The Next Generation of Surgical Navigation: Proof of Concept of an Augmented Reality System for Surgical Navigation.

Objective The aim of this work was the development of an augmented reality system including the functionality of conventional surgical navigation systems. Methods An application software for the Augmented Reality System HoloLens 2 from Microsoft was developed. It detects the position of the patient as well as position of surgical instruments in real time and displays it within the two-dimensional (2D) magnetic resonance imaging or computed tomography (CT) images. The surgical pointer instrument, including a pattern that is recognized by the HoloLens 2 sensors, was created with three-dimensional (3D) printing. The technical concept was demonstrated at a cadaver skull to identify anatomical landmarks. Results With the help of the HoloLens 2 and its sensors, the real-time position of the surgical pointer instrument could be shown. The position of the 3D-printed pointer with colored pattern could be recognized within 2D-CT images when stationary and in motion at a cadaver skull. Feasibility could be demonstrated for the clinical application of transsphenoidal pituitary surgery. Conclusion The HoloLens 2 has a high potential for use as a surgical navigation system. With subsequent studies, a further accuracy evaluation will be performed receiving valid data for comparison with conventional surgical navigation systems. In addition to transsphenoidal pituitary surgery, it could be also applied for other surgical disciplines.

Digital three-dimensional assessment of free gingival graft remodeling over 12 months.

OBJECTIVE: To digitally evaluate the three-dimensional (3D) remodelling of FGG used to treat RT2 gingival recessions and lack of keratinized tissue on mandibular incisor teeth. METHODS: Data from 45 patients included in a previous multicentric RCT were analyzed. Silicone impressions were taken before (baseline) and 3, 6 and 12 months after standardized FGG placement. Casts were scanned and images were superimposed, using digital software, to obtain measurements of estimated soft tissue thickness (eTT; 1, 3, and 5 mm apical to baseline gingival margin). In addition, soft tissue volume (STV) and creeping attachment (CA) were assessed. RESULTS: All patients exhibited postoperative eTT and STV increases, at all time points. The greatest mean thickness gain was observed at eTT3 (1.0 ± 0.4 mm) at 12 months. At 12 months, STV was 52.3 ± 21.1 mm3, without relevant changes compared to the 3- and 6-month follow-up. CA, which was observed as early as six months postoperatively, was evident in ∼85 % of teeth at 12 months. CONCLUSIONS: Application of FGG was an effective phenotype modification therapy, as shown by the significantly increased tissue thickness postoperatively. Despite the use of FGG technique not aiming for root coverage, digital 3D assessment documented the early and frequent postoperative occurrence of CA, which helped improve recession treatment outcomes. CLINICAL SIGNIFICANCE: The use of 3D assessment methodology allows precise identification of the tissue gain obtained with FGG, which, regardless of technique, results in predictable phenotype modification and frequent occurrence of creeping attachment.

Accuracy of immediate anterior implantation using static and robotic computer-assisted implant surgery: A retrospective study.

OBJECTIVES: To investigate the accuracy of immediate anterior implantation using static computer-assisted implant surgery (s-CAIS) and robotic computer-assisted implant surgery (r-CAIS). MATERIALS AND METHODS: One hundred and six implants were immediately inserted in the anterior zone of 69 patients using a freehand technique, s-CAIS or r-CAIS. Postoperative cone-beam computed tomography scans were matched with preoperative plans to evaluate the deviations between the planned and placed implant positions. RESULTS: The global coronal deviations in the freehand, s-CAIS, and r-CAIS groups were 1.29 ± 0.52 mm, 1.01 ± 0.41 mm, and 0.62 ± 0.28 mm, respectively. Significant differences were observed in the r-CAIS group compared to both the s-CAIS group and the freehand group (p < 0.05). However, no significant differences were found between the s-CAIS group and the freehand group (p > 0.05). The global apical deviations in the freehand, s-CAIS and r-CAIS groups were 1.78 ± 0.59 mm, 1.24 ± 0.52 mm and 0.65 ± 0.27 mm, respectively, while the angular deviations in the freehand, s-CAIS and r-CAIS groups were 6.46 ± 2.21°, 2.94 ± 1.71° and 1.46 ± 0.57°, respectively. Significant differences were observed in both the global apical deviations and angular deviations among the three groups (p < 0.05). CONCLUSIONS: The accuracy of immediate anterior implantation with r-CAIS was better than that with s-CAIS. This difference is attributed to better control of the coronal, vertical and axial errors during r-CAIS. CLINICAL SIGNIFICANCE: This study provides significant evidence to support the use of r-CAIS as a potential alternative in immediate anterior implantation.

Understanding the use of CATI and web-based data collection methods during the pandemic among digitally challenged groups at FQHCs: data from the All of Us Research Program.

Introduction: The All of Us Research Program (Program) is an ongoing epidemiologic cohort study focused on collecting lifestyle, health, socioeconomic, environmental, and biological data from 1 million US-based participants. The Program has a focus on enrolling populations that are underrepresented in biomedical research (UBR). Federally Qualified Health Centers (FQHCs) are a key recruitment stream of UBR participants. The Program is digital by design where participants complete surveys via web-based platform. As many FQHC participants are not digitally ready, recruitment and retention is a challenge, requiring high-touch methods. However, high-touch methods ceased as an option in March 2020 when the Program paused in-person activities because of the pandemic. In January 2021, the Program introduced Computer Assisted Telephone Interviewing (CATI) to help participants complete surveys remotely. This paper aims to understand the association between digital readiness and mode of survey completion (CATI vs. web-based platform) by participants at FQHCs. Methods: This study included 2,089 participants who completed one or more surveys via CATI and/or web-based platform between January 28, 2021 (when CATI was introduced) and January 27, 2022 (1 year since CATI introduction). Results and discussion: Results show that among the 700 not-digitally ready participants, 51% used CATI; and of the 1,053 digitally ready participants, 30% used CATI for completing retention surveys. The remaining 336 participants had "Unknown/Missing" digital readiness of which, 34% used CATI. CATI allowed survey completion over the phone with a trained staff member who entered responses on the participant's behalf. Regardless of participants' digital readiness, median time to complete retention surveys was longer with CATI compared to web. CATI resulted in fewer skipped responses than the web-based platform highlighting better data completeness. These findings demonstrate the effectiveness of using CATI for improving response rates in online surveys, especially among populations that are digitally challenged. Analyses provide insights for NIH, healthcare providers, and researchers on the adoption of virtual tools for data collection, telehealth, telemedicine, or patient portals by digitally challenged groups even when in-person assistance continues to remain as an option. It also provides insights on the investment of staff time and support required for virtual administration of tools for health data collection.

Diagnostic efficacy of virtual organ computer-assisted analysis in measuring the volume ratio of subchorionic hematoma with serum progesterone.

BACKGROUND: Subchorionic hematoma (SCH) is a common complication in early pregnancy characterized by the accumulation of blood between the uterine wall and the chorionic membrane. SCH can lead to adverse pregnancy outcomes such as miscarriage, preterm birth, and other complications. Early detection and accurate assessment of SCH are crucial for appropriate management and improved pregnancy outcomes. AIM: To evaluate the diagnostic efficacy of virtual organ computer-assisted analysis (VOCAL) in measuring the volume ratio of SCH to gestational sac (GS) combined with serum progesterone on early pregnancy outcomes in patients with SCH. METHODS: A total of 153 patients with SCH in their first-trimester pregnancies between 6 and 11 wk were enrolled. All patients were followed up until a gestational age of 20 wk. The parameters of transvaginal two-dimensional ultrasound, including the circumference of SCH (Cs), surface area of SCH (Ss), circumference of GS (Cg), and surface area of GS (Sg), and the parameters of VOCAL with transvaginal three-dimensional ultrasound, including the three-dimensional volume of SCH (3DVs) and GS (3DVg), were recorded. The size of the SCH and its ratio to the GS size (Cs/Cg, Ss/Sg, 3DVs/3DVg) were recorded and compared. RESULTS: Compared with those in the normal pregnancy group, the adverse pregnancy group had higher Cs/Cg, Ss/Sg, and 3DVs/3DVg ratios (P < 0.05). When 3DVs/3DVg was 0.220, the highest predictive performance predicted adverse pregnancy outcomes, resulting in an AUC of 0.767, and the sensitivity, specificity were 70.2%, 75% respectively. VOCAL measuring 3DVs/3DVg combined with serum progesterone gave a diagnostic AUC of 0.824 for early pregnancy outcome in SCH patients, with a high sensitivity of 82.1% and a specificity of 72.1%, which showed a significant difference between AUC. CONCLUSION: VOCAL-measured 3DVs/3DVg effectively quantifies the severity of SCH, while combined serum progesterone better predicts adverse pregnancy outcomes.

Short-Term Outcomes of Total Knee Arthroplasty Using a Conventional, Computer-Assisted, and Robotic Technique: A Pilot Clinical Trial.

Background: Total Knee Arthroplasty (TKA) is a prevalent surgical procedure for treating severe knee arthritis, aiming to alleviate pain and restore function. Recent advancements have introduced computer-assisted (CAS) and robot-assisted (RA-TKA) surgical techniques as alternatives to conventional methods, promising improved accuracy and patient outcomes. However, comprehensive comparative studies evaluating the short-term outcomes and prostheses survivorship among these techniques are limited. We hypothesized that the outcome of RA-TKA and/or CAS- TKA is advantageous in function and prosthesis survivorship compared to manually implanted prostheses. Methods: This prospective controlled study compared the short-term outcomes and prostheses survivorship following TKA using conventional, CAS, and RA-TKA techniques. One hundred seventy-eight patients requiring TKA were randomly assigned to one of the three surgical groups. The primary outcomes were knee function (KSS knee score) and functional recovery (KSS function score), which were assessed before surgery three years postoperatively. Secondary outcomes included prosthesis alignment, knee range of movements, and complication rates. Survivorship analysis was conducted using Kaplan-Meier curves, with revision surgery as the endpoint. Results: While all three groups showed significant improvements in knee function postoperatively (p < 0.001), the CAS and RA-TKA groups demonstrated superior prosthetic alignment and higher survivorship rates than the conventional group (100%, 97%, and 96%, respectively). However, although the RA-TKA group had a maximal 100% survivorship rate, its knee score was significantly lower than following CAS and conventional techniques (mean 91 ± 3SD vs. mean 93 ± 3SD, p = 0.011). Conclusion: The RA-TKA technique offers advantages over conventional and CAS methods regarding alignment accuracy and short-term survivorship of TKA prostheses. Since short-term prosthesis survivorship indicates the foreseen rates of mid- and long-term survivorship, the current data have a promising indication of the improved TKA prosthesis's long-term survivorship by implementing RA-TKA. According to the presented data, although the survival rates were 100%, 97%, and 96% in the three study groups, no clinical difference in the functional outcome was found despite the better mechanical alignment and higher survivorship in the group of patients treated by the RA-TKA.

Extravasated contrast volumetric assessment on computed tomography angiography in gastrointestinal bleeding: A useful predictor of positive angiographic findings.

BACKGROUND: Gastrointestinal bleeding (GIB) is a severe and potentially life-threatening condition, especially in cases of delayed treatment. Computed tomography angiography (CTA) plays a pivotal role in the early identification of upper and lower GIB and in the prompt treatment of the haemorrhage. AIM: To determine whether a volumetric estimation of the extravasated contrast at CTA in GIB may be a predictor of subsequent positive angiographic findings. METHODS: In this retrospective single-centre study, 35 patients (22 men; median age 69 years; range 16-92 years) admitted to our institution for active GIB detected at CTA and further submitted to catheter angiography between January 2018 and February 2022 were enrolled. Twenty-three (65.7%) patients underwent endoscopy before CTA. Bleeding volumetry was evaluated in both arterial and venous phases via a semi-automated dedicated software. Bleeding rate was obtained from volume change between the two phases and standardised for unit time. Patients were divided into two groups, according to the angiographic signs and their concordance with CTA. RESULTS: Upper bleeding accounted for 42.9% and lower GIB for 57.1%. Mean haemoglobin value at the admission was 7.7 g/dL. A concordance between positive CTA and direct angiographic bleeding signs was found in 19 (54.3%) cases. Despite no significant differences in terms of bleeding volume in the arterial phase (0.55 mL vs 0.33 mL, P = 0.35), a statistically significant volume increase in the venous phase was identified in the group of patients with positive angiography (2.06 mL vs 0.9 mL, P = 0.02). In the latter patient group, a significant increase in bleeding rate was also detected (2.18 mL/min vs 0.19 mL/min, P = 0.02). CONCLUSION: In GIB of any origin, extravasated contrast volumetric analysis at CTA could be a predictor of positive angiography and may help in avoiding further unnecessary procedures.

Femoral osteoporosis prediction model using autosegmentation and machine learning analysis with PyRadiomics on abdomen-pelvic computed tomography (CT).

Background: With the advancement of artificial intelligence technology and radiomics analysis, opportunistic prediction of osteoporosis with computed tomography (CT) is a new paradigm in osteoporosis screening. This study aimed to assess the diagnostic performance of osteoporosis prediction by the combination of autosegmentation of the proximal femur and machine learning analysis with a reference standard of dual-energy X-ray absorptiometry (DXA). Methods: Abdomen-pelvic CT scans were retrospectively analyzed from 1,122 patients who received both DXA and abdomen-pelvic computed tomography (APCT) scan from January 2018 to December 2020. The study cohort consisted of a training cohort and a temporal validation cohort. The left proximal femur was automatically segmented, and a prediction model was built by machine-learning analysis using a random forest (RF) analysis and 854 PyRadiomics features. The technical success rate of autosegmentation, diagnostic test, area under the receiver operator characteristics curve (AUC), and precision recall curve (AUC-PR) analysis were used to analyze the training and validation cohorts. Results: The osteoporosis prevalence of the training and validation cohorts was 24.5%, and 10.3%, respectively. The technical success rate of autosegmentation of the proximal femur was 99.7%. In the diagnostic test, the training and validation cohorts showed 78.4% vs. 63.3% sensitivity, 89.4% vs. 98.1% specificity. The prediction performance to identify osteoporosis within the groups used for training and validation cohort was high and the AUC and AUC-PR to forecast the occurrence of osteoporosis within the training and validation cohorts were 90.8% [95% confidence interval (CI), 88.4-93.2%] vs. 78.0% (95% CI, 76.0-79.9%) and 94.6% (95% CI, 89.3-99.8%) vs. 88.8% (95% CI, 86.2-91.5%), respectively. Conclusions: The osteoporosis prediction model using autosegmentation of proximal femur and machine-learning analysis with PyRadiomics features on APCT showed excellent diagnostic feasibility and technical success.

Causal Forest Machine Learning Analysis of Parkinson's Disease in Resting-State Functional Magnetic Resonance Imaging.

In recent years, Artificial Intelligence has been used to assist healthcare professionals in detecting and diagnosing neurodegenerative diseases. In this study, we propose a methodology to analyze functional Magnetic Resonance Imaging signals and perform classification between Parkinson's disease patients and healthy participants using Machine Learning algorithms. In addition, the proposed approach provides insights into the brain regions affected by the disease. The functional Magnetic Resonance Imaging from the PPMI and 1000-FCP datasets were pre-processed to extract time series from 200 brain regions per participant, resulting in 11,600 features. Causal Forest and Wrapper Feature Subset Selection algorithms were used for dimensionality reduction, resulting in a subset of features based on their heterogeneity and association with the disease. We utilized Logistic Regression and XGBoost algorithms to perform PD detection, achieving 97.6% accuracy, 97.5% F1 score, 97.9% precision, and 97.7%recall by analyzing sets with fewer than 300 features in a population including men and women. Finally, Multiple Correspondence Analysis was employed to visualize the relationships between brain regions and each group (women with Parkinson, female controls, men with Parkinson, male controls). Associations between the Unified Parkinson's Disease Rating Scale questionnaire results and affected brain regions in different groups were also obtained to show another use case of the methodology. This work proposes a methodology to (1) classify patients and controls with Machine Learning and Causal Forest algorithm and (2) visualize associations between brain regions and groups, providing high-accuracy classification and enhanced interpretability of the correlation between specific brain regions and the disease across different groups.

Evaluation of the resection plane three-dimensional positional accuracy using a resection guide directional guidance slot; a randomized clinical trial.

AIM: The study was performed to compare the mandibular resection guide with a directional guidance slot with the conventional guide regarding three-dimensional positional accuracy. MATERIALS AND METHODS: Twenty-six patients with lateral segmental mandibular defects were selected, and randomly allocated into two groups. All defects were managed with preoperative virtual surgical planning. Resection in the test group was conducted using a resection guide with a directional guidance slot, while a conventional resection guide design was utilized in the control group. The linear and angular deviation of the osteotomy planes was analyzed for both groups, along with the accuracy of the insertion of the reconstruction bone block in the resected defect. Data were documented, absolute deviation was calculated, statistical analysis was performed and significance was set at the 5% level. RESULTS: The cases conducted with a directional guidance templet reported a statistically significant difference when compared to the conventional edge-cutting guide regarding the linear and angular spatial osteotomy plane position (P < 0.001). The defect span analysis reported excellent levels of agreement in both groups (ICC = 1.00, ICC = 0.995), however, the difference between the groups was statistically significant (P < 0.001). CONCLUSION: The study demonstrated the enhanced positional accuracy of the resection plane and reconstruction block placement when a directional slot is incorporated in the computer-generated resection guide.

Optimization of window settings for coronary arteries assessment using spectral CT-derived virtual monoenergetic imaging.

PURPOSE: To determine the optimal window setting for virtual monoenergetic images (VMI) reconstructed from dual-layer spectral coronary computed tomography angiography (DE-CCTA) datasets. MATERIAL AND METHODS: 50 patients (30 males; mean age 61.1 ± 12.4 years who underwent DE-CCTA from May 2021 to June 2022 for suspected coronary artery disease, were retrospectively included. Image quality assessment was performed on conventional images and VMI reconstructions at 70 and 40 keV. Objective image quality was assessed using contrast-to-noise ratio (CNR). Two independent observers manually identified the best window settings (B-W/L) for VMI 70 and VMI 40 visualization. B-W/L were then normalized with aortic attenuation using linear regression analysis to obtain the optimized W/L (O-W/L) settings. Additionally, subjective image quality was evaluated using a 5-point Likert scale, and vessel diameters were measured to examine any potential impact of different W/L settings. RESULTS: VMI 40 demonstrated higher CNR values compared to conventional and VMI 70. B-W/L settings identified were 1180/280 HU for VMI 70 and 3290/900 HU for VMI 40. Subsequent linear regression analysis yielded O-W/L settings of 1155/270 HU for VMI 70 and 3230/880 HU for VMI 40. VMI 40 O-W/L received the highest scores for each parameter compared to conventional (all p < 0.0027). Using O-W/L settings for VMI 70 and VMI 40 did not result in significant differences in vessel measurements compared to conventional images. CONCLUSION: Optimization of VMI requires adjustments in W/L settings. Our results recommend W/L settings of 1155/270 HU for VMI 70 and 3230/880 HU for VMI 40.

In vitro blood sample assessment: investigating correlation of laboratory hemoglobin and spectral properties of dual-energy CT measurements (ρ/Z).

OBJECTIVES: Our study comprised a single-center retrospective in vitro correlation between spectral properties, namely ρ/Z values, derived from scanning blood samples using dual-energy computed tomography (DECT) with the corresponding laboratory hemoglobin/hematocrit (Hb/Hct) levels and assessed the potential in anemia-detection. METHODS: DECT of 813 patient blood samples from 465 women and 348 men was conducted using a standardized scan protocol. Electron density relative to water (ρ or rho), effective atomic number (Zeff), and CT attenuation (Hounsfield unit) were measured. RESULTS: Positive correlation with the Hb/Hct was shown for ρ (r-values 0.37-0.49) and attenuation (r-values 0.59-0.83) while no correlation was observed for Zeff (r-values -0.04 to 0.08). Significant differences in attenuation and ρ values were detected for blood samples with and without anemia in both genders (p value < 0.001) with area under the curve ranging from 0.7 to 0.95. Depending on the respective CT parameters, various cutoff values for CT-based anemia detection could be determined. CONCLUSION: In summary, our study investigated the correlation between DECT measurements and Hb/Hct levels, emphasizing novel aspects of ρ and Zeff values. Assuming that quantitative changes in the number of hemoglobin proteins might alter the mean Zeff values, the results of our study show that there is no measurable correlation on the atomic level using DECT. We established a positive in vitro correlation between Hb/Hct values and ρ. Nevertheless, attenuation emerged as the most strongly correlated parameter with identifiable cutoff values, highlighting its preference for CT-based anemia detection. CLINICAL RELEVANCE STATEMENT: By scanning multiple blood samples with dual-energy CT scans and comparing the measurements with standard laboratory blood tests, we were able to underscore the potential of CT-based anemia detection and its advantages in clinical practice. KEY POINTS: Prior in vivo studies have found a correlation between aortic blood pool and measured hemoglobin and hematocrit. Hemoglobin and hematocrit correlated with electron density relative to water and attenuation but not Zeff. Dual-energy CT has the potential for additional clinical benefits, such as CT-based anemia detection.

Computer-assisted patient identification tool in inborn errors of metabolism - potential for rare disease patient registry and big data analysis.

BACKGROUND: Patient registries are crucial for rare disease management. However, manual registry construction is labor-intensive and often not user-friendly. Our goal is to establish Hong Kong's first computer-assisted patient identification tool for rare diseases, starting with inborn errors of metabolism (IEM). METHODS: Patient data from 2010 to 2019 was retrieved from electronic databases. Through big data analytics, patient data were filtered based on specific IEM-related biochemical and genetic tests. Clinical notes were analyzed using a rule-based natural language processing technique called regular expression. The algorithm classified each extracted paragraph as "IEM-related" or "not IEM-related." Pathologists reviewed the paragraphs for curation, and the algorithm's performance was evaluated. RESULTS: Out of 46,419 patients with IEM-related tests, the algorithm identified 100 as "IEM-related." After pathologists' validation, 96 cases were confirmed as true IEM, with 1 uncertain case and 3 false positives. A secondary ascertainment yielded a sensitivity of 92.3% compared to our previously published IEM cohort. CONCLUSIONS: Our artificial intelligence approach provides a novel method to identify IEM patients, facilitating the creation of a centralized, computer-assisted rare disease patient registry at the local and national levels. This data can potentially be accessed by multiple stakeholders for collaborative research and to enhance healthcare management for rare diseases.