Deep learning pipeline for quality filtering of MRSI spectra.

With the rise of novel 3D magnetic resonance spectroscopy imaging (MRSI) acquisition protocols in clinical practice, which are capable of capturing a large number of spectra from a subject's brain, there is a need for an automated preprocessing pipeline that filters out bad-quality spectra and identifies contaminated but salvageable spectra prior to the metabolite quantification step. This work introduces such a pipeline based on an ensemble of deep-learning classifiers. The dataset consists of 36,338 spectra from one healthy subject and five brain tumor patients, acquired with an EPSI variant, which implemented a novel type of spectral editing named SLOtboom-Weng (SLOW) editing on a 7T MR scanner. The spectra were labeled manually by an expert into four classes of spectral quality as follows: (i) noise, (ii) spectra greatly influenced by lipid-related artifacts (deemed not to contain clinical information), (iii) spectra containing metabolic information slightly contaminated by lipid signals, and (iv) good-quality spectra. The AI model consists of three pairs of networks, each comprising a convolutional autoencoder and a multilayer perceptron network. In the classification step, the encoding half of the autoencoder is kept as a dimensionality reduction tool, while the fully connected layers are added to its output. Each of the three pairs of networks is trained on different representations of spectra (real, imaginary, or both), aiming at robust decision-making. The final class is assigned via a majority voting scheme. The F1 scores obtained on the test dataset for the four previously defined classes are 0.96, 0.93, 0.82, and 0.90, respectively. The arguably lower value of 0.82 was reached for the least represented class of spectra mildly influenced by lipids. Not only does the proposed model minimise the required user interaction, but it also greatly reduces the computation time at the metabolite quantification step (by selecting a subset of spectra worth quantifying) and enforces the display of only clinically relevant information.

icobrain ms 5.1: Combining unsupervised and supervised approaches for improving the detection of multiple sclerosis lesions.

Multiple sclerosis (MS) is a chronic autoimmune, inflammatory neurological disease of the central nervous system. Its diagnosis nowadays commonly includes performing an MRI scan, as it is the most sensitive imaging test for MS. MS plaques are commonly identified from fluid-attenuated inversion recovery (FLAIR) images as hyperintense regions that are highly varying in terms of their shapes, sizes and locations, and are routinely classified in accordance to the McDonald criteria. Recent years have seen an increase in works that aimed at development of various semi-automatic and automatic methods for detection, segmentation and classification of MS plaques. In this paper, we present an automatic combined method, based on two pipelines: a traditional unsupervised machine learning technique and a deep-learning attention-gate 3D U-net network. The deep-learning network is specifically trained to address the weaker points of the traditional approach, namely difficulties in segmenting infratentorial and juxtacortical plaques in real-world clinical MRIs. It was trained and validated on a multi-center multi-scanner dataset that contains 159 cases, each with T1 weighted (T1w) and FLAIR images, as well as manual delineations of the MS plaques, segmented and validated by a panel of raters. The detection rate was quantified using lesion-wise Dice score. A simple label fusion is implemented to combine the output segmentations of the two pipelines. This combined method improves the detection of infratentorial and juxtacortical lesions by 14% and 31% respectively, in comparison to the unsupervised machine learning pipeline that was used as a performance assessment baseline.

Improving the detection of autism spectrum disorder by combining structural and functional MRI information.

Autism Spectrum Disorder (ASD) is a brain disorder that is typically characterized by deficits in social communication and interaction, as well as restrictive and repetitive behaviors and interests. During the last years, there has been an increase in the use of magnetic resonance imaging (MRI) to help in the detection of common patterns in autism subjects versus typical controls for classification purposes. In this work, we propose a method for the classification of ASD patients versus control subjects using both functional and structural MRI information. Functional connectivity patterns among brain regions, together with volumetric correspondences of gray matter volumes among cortical parcels are used as features for functional and structural processing pipelines, respectively. The classification network is a combination of stacked autoencoders trained in an unsupervised manner and multilayer perceptrons trained in a supervised manner. Quantitative analysis is performed on 817 cases from the multisite international Autism Brain Imaging Data Exchange I (ABIDE I) dataset, consisting of 368 ASD patients and 449 control subjects and obtaining a classification accuracy of 85.06 ± 3.52% when using an ensemble of classifiers. Merging information from functional and structural sources significantly outperforms the implemented individual pipelines.

Primary malignant fibrous histiocytoma of the spleen: recurrence eight years after splenectomy--report of a case and literature review.

Primary intraabdominal malignant mesenchymal tumors are very rare. There are just few cases of intraabdominal visceral malignant fibrous histiocytoma in the literature. We report a case of primary malignant fibrous histiocytoma of the spleen in a 57-year-old man, with a recurrence eight years after the splenectomy. After the initial surgery the patient was without complaints, and refused to receive chemotherapy or radiotherapy. Eight years after the surgery the patient reported due to general weakness and malaise when the diagnosis of disease relapse was established. Radical surgery was performed although the tumor involved large curvature of the stomach, left crus of the diaphragm, splenic flexure of the colon and tail of pancreas. Four months after the surgery patient died. To the best of our knowledge, to date, only 18 cases have been reported in the literature, describing malignant fibrous histiocytoma of the spleen.

Genu recurvatum as a complication after total knee arthroplasty.

A 73-year-old female patient underwent total knee arthroplasty for arthrosis. After surgery, recurvatum instability over 55 degrees occurred. Two years later, primary endoprosthesis was removed and a revision endoprosthesis implanted. However, complete deformity occurred again. Two years later, the thickest revision polyethylene tibial implant was implanted on the same endoprosthesis. In spite of using knee orthosis, recurvatum deformity returned. Seven months later, new revision rotating hinge prosthesis was implanted. This procedure seems to have solved the problem. The knee deformity in this specific case was larger than the deformity in cases reported so far, and it was solved after three surgical procedures. Rotating hinge prosthesis seems to be the method of choice for immediate repair of recurvatum instability after total knee arthroplasty.

New paradigm in training of undergraduate clinical skills: the NEPTUNE-CS project at the Split University School of Medicine.

Clinical skills' training is arguably the weakest point in medical schools' curriculum. This study briefly describes how we at the Split University School of Medicine cope with this problem. We consider that, over the last decades, a considerable advancement in teaching methodologies, tools, and assessment of students has been made. However, there are many unresolved issues, most notably: (i) the institutional value system, impeding the motivation of the teaching staff; (ii) lack of a strong mentoring system; (iii) organization, timing, and placement of training in the curriculum; (iv) lack of publications pertinent to training; and (v) unwillingness of patients to participate in student training. To improve the existing training models we suggest increased institutional awareness of obstacles, as well as willingness to develop mechanisms for increasing the motivation of faculty. It is necessary to introduce changes in the structure and timing of training and to complement it with a catalog, practicum, and portfolio of clinical skills. At Split University School of Medicine, we developed a new paradigm aimed to improve the teaching of clinical skills called "Neptune-CSS," which stands for New Paradigm in Training of Undergraduate Clinical Skills in Split.

Spinal cord stimulation for chronic visceral pain secondary to chronic non-alcoholic pancreatitis.

Spinal cord stimulation (SCS) suppresses visceral response to colon distension in an animal model. In humans, it may be an effective therapy for chronic pain of pelvic origin, irritable bowel syndrome, and persistent unspecified abdominal pain. Described here is the case of SCS for 38-year-old woman with visceral pain secondary to chronic pancreatitis. Previous therapies included numerous endoscopic retrograde cholangiopancreatographies, multiple pancreatic duct stenting, chemical and surgical sympathectomies with short-lasting pain relief. After the initial evaluation, the patient underwent retrograde epidural differential block to determine possible source of pain. Delay in pain recurrence after block suggested that the origin of her pain was visceral. After the psychologic evaluation, the patient underwent SCS trial over 14 days. She had 2 trial leads placed epidurally via T9-T10 paramedian entry with the tips of both leads positioned at T6 vertebral body. During the trial, visual analog scale pain score decreased from 8 to 1 cm, Pain Disability Index from 62 to 14, and opioid use from 150 to 0 mg of morphine sulfate equivalent a day. After the completion of successful SCS trial, she was implanted with dual octrode leads and rechargeable pulse generator. Median pain scores decreased from 8 to 1 at 3 months after the implant. Pain Disability Index changed from 62 to 15. Opiate use decreased to none. It seems that SCS may have a significant therapeutic potential for the treatment of visceral pain secondary to chronic pancreatitis.

Age-related attenuation of isoflurane preconditioning in human atrial cardiomyocytes: roles for mitochondrial respiration and sarcolemmal adenosine triphosphate-sensitive potassium channel activity.

BACKGROUND: Clinical trials suggest that anesthetic-induced preconditioning (APC) produces cardioprotection in humans, but the mechanisms of APC and significance of aging for APC in humans are not well understood. Here, the impact of age on the role of two major effectors of APC, mitochondria and sarcolemmal adenosine triphosphate-sensitive potassium (sarcKATP) channels, in preconditioning of the human atrial myocardium were investigated. METHODS: Right atrial appendages were obtained from adult patients undergoing cardiac surgery and assigned to mid-aged (MA) and old-aged (OA) groups. APC was induced by isoflurane in isolated myocardium and isolated cardiomyocytes. Mitochondrial oxygen consumption measurements, myocyte survival testing, and patch clamp techniques were used to investigate mitochondrial respiratory function and sarcKATP channel activity. RESULTS: After in vitro APC with isoflurane, the respiratory function of isolated mitochondria was better preserved after hypoxia-reoxygenation stress in MA than in OA. In isolated intact myocytes, APC significantly decreased oxidative stress-induced cell death in MA but not in OA, and isoflurane protection from cell death was attenuated by the sarcKATP channel inhibitor HMR-1098. Further, the properties of single sarcKATP channels were similar in MA and OA, and isoflurane sensitivity of pinacidil-activated whole cell KATP current was no different between MA and OA myocytes. CONCLUSION: Anesthetic-induced preconditioning with isoflurane decreases stress-induced cell death and preserves mitochondrial respiratory function to a greater degree in MA than in OA myocytes; however, sarcKATP channel activity is not differentially affected by isoflurane. Therefore, effectiveness of APC in humans may decrease with advancing age partly because of altered mitochondrial function of myocardial cells.

Application of novel lossless compression of medical images using prediction and contextual error modeling.

Conduction of tele-3D-computer assisted operations as well as other telemedicine procedures often requires highest possible quality of transmitted medical images and video. Unfortunately, those data types are always associated with high telecommunication and storage costs that sometimes prevent more frequent usage of such procedures. We present a novel algorithm for lossless compression of medical images that is extremely helpful in reducing the telecommunication and storage costs. The algorithm models the image properties around the current, unknown pixel and adjusts itself to the local image region. The main contribution of this work is the enhancement of the well known approach of predictor blends through highly adaptive determination of blending context on a pixel-by-pixel basis using classification technique. We show that this approach is well suited for medical image data compression. Results obtained with the proposed compression method on medical images are very encouraging, beating several well known lossless compression methods. The predictor proposed can also be used in other image processing applications such as segmentation and extraction of image regions.

Comparison of on-demand vs planned relaparotomy for treatment of severe intra-abdominal infections.

AIM: To compare morbidity and mortality of patients with severe intra-abdominal infections after two types of surgical treatment, on-demand ("wait and see") relaparotomy and modified planned relaparotomy. METHODS: We prospectively analyzed the outcomes of 65 patients with severe peritonitis surgically treated in two Croatian hospitals. In one hospital, 34 patients were treated on-demand, and in another hospital 31 patients were treated by planned relaparotomy. We compared severe postoperative complications, mortality, and length of hospital stay in the two groups of patients. RESULTS: Severity of patient's disease, as measured from preoperative group-average Acute Physiology and Chronic Health Evaluation (APACHE) II scores, was comparable in both on-demand and planned relaparotomy groups. The mortality rate was higher in patients operated on-demand (59% vs 29%, P=0.024). In nonadjusted model, the relative risk of dying was 2.5-fold higher for patients treated by on-demand operation in comparison with planned relaparatomy (P=0.030). However, after the adjustment of the survival data for individual patient's sex and APACHE II scores, the difference in the relative risk became non-significant (P=0.178). The patients who died had higher APACHE II scores (26.1+/-8.9 vs 19.7+/-5.9, P=0.009). Relative risk of dying per 5-point increase in APACHE II score was 1.24 (95% confidence interval, 1.01-1.51; P=0.039), irrespective of the surgical technique. CONCLUSIONS: Patients with severe peritonitis treated with planned relaparotomy seemed to have lower mortality. However, the relative risk of dying was not statistically different between the on-demand and planned relaparotomy groups after adjustment for preoperative APACHE II scores. The severity of disease rather than surgical approach plays more important role in survival of these patients.