Machine learning-based radiomic analysis and growth visualization for ablation site recurrence diagnosis in follow-up CT.

OBJECTIVES: Detecting ablation site recurrence (ASR) after thermal ablation remains a challenge for radiologists due to the similarity between tumor recurrence and post-ablative changes. Radiomic analysis and machine learning methods may show additional value in addressing this challenge. The present study primarily sought to determine the efficacy of radiomic analysis in detecting ASR on follow-up computed tomography (CT) scans. The second aim was to develop a visualization tool capable of emphasizing regions of ASR between follow-up scans in individual patients. MATERIALS AND METHODS: Lasso regression and Extreme Gradient Boosting (XGBoost) classifiers were employed for modeling radiomic features extracted from regions of interest delineated by two radiologists. A leave-one-out test (LOOT) was utilized for performance evaluation. A visualization method, creating difference heatmaps (diff-maps) between two follow-up scans, was developed to emphasize regions of growth and thereby highlighting potential ASR. RESULTS: A total of 55 patients, including 20 with and 35 without ASR, were included in the radiomic analysis. The best performing model was achieved by Lasso regression tested with the LOOT approach, reaching an area under the curve (AUC) of 0.97 and an accuracy of 92.73%. The XGBoost classifier demonstrated better performance when trained with all extracted radiomic features than without feature selection, achieving an AUC of 0.93 and an accuracy of 89.09%. The diff-maps correctly highlighted post-ablative liver tumor recurrence in all patients. CONCLUSIONS: Machine learning-based radiomic analysis and growth visualization proved effective in detecting ablation site recurrence on follow-up CT scans.

Comparison of Two 2.45 GHz Microwave Ablation Devices with Respect to Ablation Zone Volume in Relation to Applied Energy in Patients with Malignant Liver Tumours.

Purpose: (i) to compare two 2.45 GHz MWA devices with respect to AZV in relation to the applied energy after MWA in patients with hepatocellular carcinoma (HCC) or colorectal liver metastasis (CRLM) and (ii) to identify potential confounders for this relationship. Methods: In total, 102 tumours, 65 CRLM and 37 HCC were included in this retrospective analysis. Tumours were treated with Emprint (n = 71) or Neuwave (n = 31) MWA devices. Ablation treatment setting were recorded and applied energy was calculated. AZV and tumour volumes were segmented on the contrast-enhanced CT scans obtained 1 week after treatment. The AZV to applied energy R(AZV:E) ratios were calculated for each tumour treatment and compared between both MWA devices and tumour types. Results: R(AZV:E)EMPRINT was 0.41 and R(AZV:E)NEUWAVE was 0.81, p < 0.001. Moderate correlation between AZV and applied energy was found for Emprint (r = 0.57, R2 = 0.32, p < 0.001) and strong correlation was found for Neuwave (r = 0.78, R2 = 0.61, p < 0.001). R(AZV:E)CRLM was 0.45 and R(AZV:E)HCC was 0.52, p = 0.270. Conclusion: This study confirms the unpredictability of AZVs based on the applied output energy for HCC and CRLM. No significant differences in R(AZV:E) were observed between CRLM and HCC. Significantly lower R(AZV:E) was found for Emprint devices compared to Neuwave; however, reflected energy due to cable and antenna design remains unclear and might contribute to these differences.

Outcomes after primary and repeat thermal ablation of hepatocellular carcinoma with or without liver transplantation.

OBJECTIVES: Thermal ablation (TA) is an established treatment for early HCC. There is a lack of data on the efficacy of repeated TA for recurrent HCC, resulting in uncertainty whether good oncologic outcomes can be obtained without performing orthotopic liver transplantation (OLTx). This study analyses outcomes after TA, with a special focus on repeat TA for recurrent HCC, either as a stand-alone therapy, or in relationship with OLTx. METHODS: Data from a prospectively registered database on interventions for HCC in a tertiary hepatobiliary centre was completed with follow-up until December 2020. Outcomes studied were rate of recurrence after primary TA and after its repeat interventions, the occurrence of untreatable recurrence, OS and DSS after primary and repeat TA, and complications after TA. In cohorts matched for confounders, OSS and DSS were compared after TA with and without the intention to perform OLTx. RESULTS: After TA, 100 patients (56·8%) developed recurrent HCC, of whom 76 (76·0%) underwent up to four repeat interventions. During follow-up, 76·7% of patients never developed a recurrence unamenable to repeat TA or OLTx. OS was comparable after primary TA and repeat TA. In matched cohorts, OS and DSS were comparable after TA with and without the intention to perform OLTx. CONCLUSIONS: We found TA to be an effective and repeatable therapy for primary and recurrent HCC. Most recurrences can be treated with curative intent. There are patients who do well with TA alone without ever undergoing OLTx. KEY POINTS: • Recurrent HCC after primary TA can often be treated effectively with repeat TA. Survival after repeat TA is comparable to primary TA. • In matched cohorts, outcomes after TA with and without subsequent waitlisting for OLTx are comparable. • There are patients who do well for many years with primary and repeat TA alone; some despite multiple recurrences.

The Role of CT in Assessment of Extraregional Lymph Node Involvement in Pancreatic and Periampullary Cancer: A Diagnostic Accuracy Study.

Purpose: To investigate the diagnostic accuracy of CT in assessing extraregional lymph node metastases in pancreatic head and periampullary cancer. Materials and Methods: This prospective observational cohort study was performed at two tertiary hepatopancreatobiliary (HPB) referral centers between March 2013 and December 2014. Patients undergoing pancreatoduodenectomy or bypass surgery with or without palliative radiofrequency ablation were included. Extraregional lymph node involvement was defined as positive lymph nodes in the aortocaval window. Two expert HPB radiologists assessed aortocaval lymph nodes at preoperative CT according to a standardized protocol. All tissue from the aortocaval window was collected intraoperatively. Positive histopathologic finding was the reference standard. Analysis of predictive values and diagnostic accuracy was performed. Results: A total of 198 consecutive patients (mean age, 66 years; range, 39-86 years; 105 men) with pancreatic head or periampullary carcinoma were included. In 70% of patients, a pancreatoduodenectomy was performed, 4% underwent total pancreatectomy, 4% underwent radiofrequency ablation, and 22% underwent bypass surgery. Forty-four patients (22%) had histologically positive aortocaval lymph nodes. Negative predictive value of CT in assessing aortocaval lymph nodes was 80% for both observers, and positive predictive value was 31%-33%. Overall diagnostic accuracy was 69%-70%. Conclusion: CT has a low diagnostic accuracy in assessing extraregional lymph node metastases in patients suspected of having pancreatic or periampullary cancer.Keywords: CT, Abdomen/GI, Pancreas, Oncology© RSNA, 2021.

Clinical and Radiologic Predictors of Parastomal Hernia Development After End Colostomy.

OBJECTIVE. Parastomal hernia (PSH) is a common complication that can occur after end colostomy and may result in considerable morbidity. To select the best candidates for prophylactic measures, knowledge of preoperative PSH predictors is important. This study aimed to determine the value of clinical parameters, preoperative CT-based body metrics, and size of the abdominal wall defect created during end colostomy and measured at postoperative CT for predicting PSH development. MATERIALS AND METHODS. Sixty-five patients who underwent permanent end colostomy with at least 1 year of follow-up were included. On preoperative CT, waist circumference, abdominal wall and psoas muscle indexes, rectus abdominis muscle diameter and diastasis, intra- and extraabdominal fat mass, and presence of other hernias were assessed. On postoperative CT, size of the abdominal wall defect and the presence of PSH were determined. To identify independent predictors of PSH development, univariate analysis with the Kaplan-Meier method and multivariate Cox regression analysis were performed. RESULTS. PSH developed after surgery in 30 patients (46%). Three independent risk factors were identified: chronic obstructive pulmonary disease (COPD) as a comorbidity (hazard ratio [HR], 6.4; 95% CI, 1.9-22.0; p = 0.003), operation time longer than 395 minutes (HR, 3.9; 95% CI, 1.5-10.0; p = 0.005), and maximum aperture diameter of more than 34 mm (HR, 5.2; 95% CI, 2.1-12.7; p < 0.001). PSH developed in all nine patients with a maximum abdominal wall defect diameter of more than 50 mm at the ostomy site. CONCLUSION. COPD, longer operation time, and larger abdominal wall defect at the colostomy site can predict PSH development. Intraoperative creation of an abdominal wall ostomy opening that is more than 34 mm in diameter should be avoided.

Body Composition of Infants With Biliary Atresia: Anthropometric Measurements and Computed Tomography-based Body Metrics.

OBJECTIVES: Biliary atresia (BA) causes neonatal cholestasis that requires hepatoportoenterostomy or liver transplantation (LT) for long-term survival. Nutritional optimization is necessary as sarcopenia and sarcopenic obesity have been associated with adverse clinical outcome. Currently, mid upper arm circumference (MUAC) is considered the most accurate indicator. The aim of the study was to determine computed tomography (CT)-based body metrics in infants with BA and to evaluate its correlation with MUAC. METHODS: We retrospectively analyzed all BA infants below 2 years of age who underwent CT as part of LT screening at our hospital between 2006 and 2019. Measured variables were indexed with length and included: MUAC, total psoas muscle surface area (tPMSA), cross-sectional skeletal muscle area (CSMA), and total abdominal fat area. Intraclass correlation coefficients and Pearson coefficients were calculated. CSMA-to-abdominal fat area ratio was divided in quartiles, the lowest quartile group was considered sarcopenic obese. RESULTS: Eighty infants with a median age of 4.6 months at LT screening were included. Intraclass correlation coefficients were: tPMSA = 0.94, CSMA = 0.92, and total abdominal fat area = 0.99. Correlation between MUAC z-score and indices of tPMSA, CSMA, and total abdominal fat area were r = 0.02, r = 0.06, and r = 0.43, respectively. The cut-off for sarcopenic obesity was CSMA-to-abdominal fat area ratio below 0.93. CONCLUSIONS: In BA infants, it is possible to determine CT-based body metrics during LT screening with very strong interobserver agreement. Poor correlation between CT-based body metrics and MUAC suggests that CT-based body metrics provide additional information on body composition in BA infants, such as relative muscle mass.

Routine Chest Computed Tomography for Staging of Pancreatic Head Carcinoma.

OBJECTIVES: The diagnostic value of routine chest computed tomography (CT) in addition to abdominal CT in workup for pancreatic head carcinoma is unclear. The aim of this study was to determine if routine chest CT revealed significant lesions that altered the management of patients with suspected pancreatic head carcinoma. METHODS: All Dutch pancreatic cancer centers were surveyed on the use of chest CT in preoperative staging. In addition, a single-center retrospective cohort study was performed including all patients referred with suspected pancreatic head malignancy without chest CT between 2005 and 2016. The primary end point was the proportion of patients in which chest CT revealed clinically significant lesions, leading to a change in management. RESULTS: In 7 of 18 Dutch pancreatic cancer centers (39%), a preoperative chest CT is not routinely performed. In the study cohort, 170 of 848 patients (20%) were referred without chest CT and underwent one by local protocol. Chest CT revealed new suspicious lesions in 17 patients (10%), of whom 6 had metastatic disease (3.5%). CONCLUSIONS: Routine use of chest CT in diagnostic workup for pancreatic head carcinoma reveals clinically significant lesions in 10% of patients, being metastases in up to 4%.

Robotic versus Freehand Needle Positioning in CT-guided Ablation of Liver Tumors: A Randomized Controlled Trial.

Purpose To compare the accuracy of freehand versus robotic antenna placement in CT-guided microwave ablation (MWA) of liver tumors. Materials and Methods This study was conducted as a prospective single-center nonblinded randomized controlled trial (Netherlands Trial Registry, NTR6023). Eligible study participants had undergone clinically indicated CT-guided MWA of liver tumors and were able to receive a CT contrast agent. Randomization was performed per tumor after identification on contrast material-enhanced CT images. The primary outcome was the number of antenna repositionings, which was compared by using the Mann-Whitney U test. Secondary outcomes were lateral targeting error stratified by in-plane and out-of-plane targets and targeting time. Results Between February 14 and November 12, 2017, 31 participants with a mean age of 63 years (range, 25-88 years) were included: 17 women (mean age, 57 years; range, 25-77 years) and 14 men (mean age, 70 years; range, 52-88 years). The freehand study arm consisted of 19 participants, while the robotic study arm consisted of 18 participants; six participants with multiple tumors were included in both arms. Forty-seven tumors were assessed; five tumors were excluded from the analysis because of technical limitations. In the robotic arm, no antenna repositioning was required. In the freehand arm, a median of one repositioning was required (range, zero to seven repositionings; P < .001). For out-of-plane targets, lateral targeting error was 10.1 mm ± 4.0 and 5.9 mm ± 2.9 (P = .007) for freehand and robotic procedures, respectively, and for in-plane targets, lateral targeting error was 6.2 mm ± 2.7 and 7.7 mm ± 5.9, respectively (P = .51). Mean targeting time was 19 minutes (range, 8-55 minutes) and 36 minutes (range, 3-70 minutes; P = .001) for freehand and robotic procedures, respectively. Conclusion Robotic antenna guidance reduces the need for antenna repositioning in microwave ablation to accurately target liver tumors and increases accuracy for out-of-plane targets. However, targeting time was greater with robotic guidance than with freehand targeting. © RSNA, 2019.

The Additional Value of Laparoscopic Ultrasound to Staging Laparoscopy in Patients with Suspected Pancreatic Head Cancer.

BACKGROUND: This study aimed to evaluate the additional value of laparoscopic ultrasound (LUS) to staging laparoscopy (SL) for detecting occult liver metastases in patients with potentially resectable pancreatic head cancer. METHODS: A retrospective cohort study was performed including all patients who underwent SL and LUS between 2005 and 2016. LUS was performed during SL to detect liver metastases not found by preoperative imaging or visual inspection of the liver. RESULTS: Out of 197 patients, visual inspection during SL detected distant metastases in 29 (14.7%) patients. LUS was performed in 127 patients, revealing 3 additional liver metastases. The proportion of patients with unresectable disease after SL and negative LUS was 32.3%, which was similar to 36.6% of patients with unresectable disease after SL without LUS (difference 4.3%; 95% CI - 13-23%; P = 0.61). Sensitivity, specificity, and positive and negative predictive values of LUS to detect liver metastases were 30, 100, 100, and 94%, respectively. The proportion of patients with distant metastases diagnosed at SL significantly increased over time (P = 0.031). CONCLUSION: The routine use of LUS during SL for patients with potentially resectable pancreatic head cancer cannot be recommended. Imaging should be repeated when significant delay occurs between index CT and the scheduled surgery.