Metrics reloaded: recommendations for image analysis validation.

Increasing evidence shows that flaws in machine learning (ML) algorithm validation are an underestimated global problem. In biomedical image analysis, chosen performance metrics often do not reflect the domain interest, and thus fail to adequately measure scientific progress and hinder translation of ML techniques into practice. To overcome this, we created Metrics Reloaded, a comprehensive framework guiding researchers in the problem-aware selection of metrics. Developed by a large international consortium in a multistage Delphi process, it is based on the novel concept of a problem fingerprint-a structured representation of the given problem that captures all aspects that are relevant for metric selection, from the domain interest to the properties of the target structure(s), dataset and algorithm output. On the basis of the problem fingerprint, users are guided through the process of choosing and applying appropriate validation metrics while being made aware of potential pitfalls. Metrics Reloaded targets image analysis problems that can be interpreted as classification tasks at image, object or pixel level, namely image-level classification, object detection, semantic segmentation and instance segmentation tasks. To improve the user experience, we implemented the framework in the Metrics Reloaded online tool. Following the convergence of ML methodology across application domains, Metrics Reloaded fosters the convergence of validation methodology. Its applicability is demonstrated for various biomedical use cases.

Combination of computed tomography angiography with coronary artery calcium score for improved diagnosis of coronary artery disease: a collaborative meta-analysis of stable chest pain patients referred for invasive coronary angiography.

OBJECTIVES: Coronary computed tomography angiography (CCTA) has higher diagnostic accuracy than coronary artery calcium (CAC) score for detecting obstructive coronary artery disease (CAD) in patients with stable chest pain, while the added diagnostic value of combining CCTA with CAC is unknown. We investigated whether combining coronary CCTA with CAC score can improve the diagnosis of obstructive CAD compared with CCTA alone. METHODS: A total of 2315 patients (858 women, 37%) aged 61.1 ± 10.2 from 29 original studies were included to build two CAD prediction models based on either CCTA alone or CCTA combined with the CAC score. CAD was defined as at least 50% coronary diameter stenosis on invasive coronary angiography. Models were built by using generalized linear mixed-effects models with a random intercept set for the original study. The two CAD prediction models were compared by the likelihood ratio test, while their diagnostic performance was compared using the area under the receiver-operating-characteristic curve (AUC). Net benefit (benefit of true positive versus harm of false positive) was assessed by decision curve analysis. RESULTS: CAD prevalence was 43.5% (1007/2315). Combining CCTA with CAC improved CAD diagnosis compared with CCTA alone (AUC: 87% [95% CI: 86 to 89%] vs. 80% [95% CI: 78 to 82%]; p < 0.001), likelihood ratio test 236.3, df: 1, p < 0.001, showing a higher net benefit across almost all threshold probabilities. CONCLUSION: Adding the CAC score to CCTA findings in patients with stable chest pain improves the diagnostic performance in detecting CAD and the net benefit compared with CCTA alone. CLINICAL RELEVANCE STATEMENT: CAC scoring CT performed before coronary CTA and included in the diagnostic model can improve obstructive CAD diagnosis, especially when CCTA is non-diagnostic. KEY POINTS: • The combination of coronary artery calcium with coronary computed tomography angiography showed significantly higher AUC (87%, 95% confidence interval [CI]: 86 to 89%) for diagnosis of coronary artery disease compared to coronary computed tomography angiography alone (80%, 95% CI: 78 to 82%, p < 0.001). • Diagnostic improvement was mostly seen in patients with non-diagnostic C. • The improvement in diagnostic performance and the net benefit was consistent across age groups, chest pain types, and genders.

Impact of an AI software on the diagnostic performance and reading time for the detection of cerebral aneurysms on time of flight MR-angiography.

PURPOSE: To evaluate the impact of an AI-based software trained to detect cerebral aneurysms on TOF-MRA on the diagnostic performance and reading times across readers with varying experience levels. METHODS: One hundred eighty-six MRI studies were reviewed by six readers to detect cerebral aneurysms. Initially, readings were assisted by the CNN-based software mdbrain. After 6 weeks, a second reading was conducted without software assistance. The results were compared to the consensus reading of two neuroradiological specialists and sensitivity (lesion and patient level), specificity (patient level), and false positives per case were calculated for the group of all readers, for the subgroup of physicians, and for each individual reader. Also, reading times for each reader were measured. RESULTS: The dataset contained 54 aneurysms. The readers had no experience (three medical students), 2 years experience (resident in neuroradiology), 6 years experience (radiologist), and 12 years (neuroradiologist). Significant improvements of overall specificity and the overall number of false positives per case were observed in the reading with AI support. For the physicians, we found significant improvements of sensitivity on lesion and patient level and false positives per case. Four readers experienced reduced reading times with the software, while two encountered increased times. CONCLUSION: In the reading with the AI-based software, we observed significant improvements in terms of specificity and false positives per case for the group of all readers and significant improvements of sensitivity and false positives per case for the physicians. Further studies are needed to investigate the effects of the AI-based software in a prospective setting.

The RNA binding protein human antigen R is a gatekeeper of liver homeostasis.

BACKGROUND AND AIMS: NAFLD is initiated by steatosis and can progress through fibrosis and cirrhosis to HCC. The RNA binding protein human antigen R (HuR) controls RNAs at the posttranscriptional level; hepatocyte HuR has been implicated in the regulation of diet-induced hepatic steatosis. The present study aimed to understand the role of hepatocyte HuR in NAFLD development and progression to fibrosis and HCC. APPROACH AND RESULTS: Hepatocyte-specific, HuR-deficient mice and control HuR-sufficient mice were fed either a normal diet or an NAFLD-inducing diet. Hepatic lipid accumulation, inflammation, fibrosis, and HCC development were studied by histology, flow cytometry, quantitative PCR, and RNA sequencing. The liver lipidome was characterized by lipidomics analysis, and the HuR-RNA interactions in the liver were mapped by RNA immunoprecipitation sequencing. Hepatocyte-specific, HuR-deficient mice displayed spontaneous hepatic steatosis and fibrosis predisposition compared to control HuR-sufficient mice. On an NAFLD-inducing diet, hepatocyte-specific HuR deficiency resulted in exacerbated inflammation, fibrosis, and HCC-like tumor development. A multi-omic approach, including lipidomics, transcriptomics, and RNA immunoprecipitation sequencing revealed that HuR orchestrates a protective network of hepatic-metabolic and lipid homeostasis-maintaining pathways. Consistently, HuR-deficient livers accumulated, already at steady state, a triglyceride signature resembling that of NAFLD livers. Moreover, up-regulation of secreted phosphoprotein 1 expression mediated, at least partially, fibrosis development in hepatocyte-specific HuR deficiency on an NAFLD-inducing diet, as shown by experiments using antibody blockade of osteopontin. CONCLUSIONS: HuR is a gatekeeper of liver homeostasis, preventing NAFLD-related fibrosis and HCC, suggesting that the HuR-dependent network could be exploited therapeutically.

Challenges and opportunities for bioimage analysis core-facilities.

Recent advances in microscopy imaging and image analysis motivate more and more institutes worldwide to establish dedicated core-facilities for bioimage analysis. To maximise the benefits research groups at these institutes gain from their core-facilities, they should be established to fit well into their respective environment. In this article, we introduce common collaborator requests and corresponding potential services core-facilities can offer. We also discuss potential competing interests between the targeted missions and implementations of services to guide decision makers and core-facility founders to circumvent common pitfalls.

New Crystalline Salts of Nicotinamide Riboside as Food Additives.

NR+ is a highly effective vitamin B3 type supplement due to its unique ability to replenish NAD+ levels. While NR+ chloride is already on the market as a nutritional supplement, its synthesis is challenging, expensive, and low yielding, making it cumbersome for large-scale industrial production. Here we report the novel crystalline NR+ salts, d/l/dl-hydrogen tartrate and d/l/dl-hydrogen malate. Their high-yielding, one-pot manufacture does not require specific equipment and is suitable for multi-ton scale production. These new NR+ salts seem ideal for nutritional applications due to their bio-equivalence compared to the approved NR+ chloride. In addition, the crystal structures of all stereoisomers of NR+ hydrogen tartrate and NR+ hydrogen malate and a comparison to the known NR+ halogenides are presented.

Patient Satisfaction and Clinical Complications after Endoscopic or Surgical Saphenous Vein Harvesting.

BACKGROUND: Endoscopic saphenous vein harvesting (EVH) for coronary artery bypass grafting (CABG) has been developed to reduce leg wound problems. This study was undertaken to evaluate postoperative complications and patient's subjective satisfaction comparing EVH and surgical vein harvesting (SVH). METHODS: From January to December 2017, patients undergoing elective cardiac surgery (CABG, CABG + Valve repair or replacement) underwent saphenous vein graft harvesting either by EVH (n = 136) or SVH (n = 104). Clinical follow-up was scheduled for day 7 and > 45 days after surgery. Primary end points were divided into two subgroups. The first one included postoperative extent of subjective pain and satisfaction with the cosmetic results described by the patients themselves, while the second subgroup included objective postoperative complications including wound healing disturbances, hematoma, and neuropathy. Secondary end point was length of hospital stay. RESULTS: At 7 days follow-up, EVH patients were more satisfied with the cosmetic results than those of the SVH group (p < 0.001) and expressed a significant tendency toward lower subjective pain compared with the SVH patients (p < 0.001), exhibited significantly lower cellulitis (p-0.002), neuropathy (p-0.005), and superficial wound healing disturbance (p-0.007). During further follow-up at > 45 days, patients with EVH were still more satisfied with the cosmetic results (p < 0.001) and expressed lower subjective pain (p < 0.001), while the other objective wound parameters did not show significant differences between both groups. Mean length of hospital stay of EVH patients was 0.7 days less compared with SVH patients. CONCLUSIONS: Our findings demonstrate the noninferiority of EVH in the short term and in the early medium term.

Clinical Short-Term Outcome and Hemodynamic Comparison of Six Contemporary Bovine Aortic Valve Prostheses.

BACKGROUND: Conventional stented valves (CV) remain gold standard for aortic valve disease. Bovine prostheses have been improved and rapid deployment valves (RDV) have arrived in the recent decade. We compare clinical and hemodynamic short-term outcome of six bovine valves. METHODS: We retrospectively evaluated 829 consecutive patients (all-comers) receiving bovine aortic valve replacement (AVR). Four CV from different manufacturers (Mitroflow, Crown, Perimount, Trifecta) and two RDV (Perceval, Intuity) were compared in terms of pre-, intra-, and postprocedural data. A risk model for mortality was created. RESULTS: All valves reduced gradients. From 23 mm, all CV showed acceptable gradients. Twenty-one millimeter Mitroflow/Perceval and 19 mm Crown showed above-average gradients. As baseline data differed, we performed propensity matching between aggregated isolated CV and RDV groups. Cardiopulmonary bypass (CPB), clamp, and surgery times were shorter with RDV (87.4 ± 34.0 min vs 111.0 ± 34.2, 54.3 ± 21.1 vs 74.9 ± 20.4, 155.2 ± 42.9 vs 178.0 ± 46.8, p < 0.001). New pacemaker rate (10.1 vs 1.3%, p = 0.016) and the tendency toward neurologic events (8.9 vs 2.5%, p = 0.086) were higher using RDV, induced mainly by the Perceval. Early mortality was equal (2.5 vs 1.3%, p = 0.560). Revision for bleeding, dialysis, blood products, length-of-stay, gradients, and regurgitation was also equal. Risk analysis showed that low valve size, low ejection fraction, endocarditis, administration of red cells, and prolonged CPB time were predictors of elevated mortality. CONCLUSION: Isolated bovine AVR has low mortality. Valves ≥ 23 mm show comparable gradients while the valve model matters < 23 mm. RDV should be used with care. Procedure-related times are shorter than those of CV but pacemaker implantation and neurologic events are more frequent (Perceval). Early mortality is low and valve performance comparable to CV.

How Do Transcatheter Heart Valves Fit in Mitral Annuloplasty Rings and Which Combination Can be Recommended?

BACKGROUND: Transcatheter heart valve (THV) as valve-in-ring is increasingly used in the mitral position. Semi-rigid rings may serve as a more appropriate scaffold for proper anchoring of a THV as they may change from their oval to a round shape thereby fitting to the implanted THV. METHODS: One rigid and five semi-rigid rings of four manufacturers, Edwards Physio I and II, Sorin 3D Memo, Medtronic Simulus, and St. Jude Medical (SJM) Saddle and SJM Sequin, with sizes 28 to 36 mm and Edwards Sapien III THV 23, 26, and 29 mm were used. Preevaluation comprised insertion/inflation of the THV into the ring and visual inspection for the paravalvular gap ≥ 4 mm2. Only valves not showing paravalvular gap were then submitted to hemodynamic evaluation with a pulse duplicator. Cusp movement was assessed with a high-speed-camera. Mean transvalvular gradients (TVGs) were measured. RESULTS: SJM Saddle ring of all sizes and SJM Sequin ring 34 showed marked gaps combined with all THV sizes, thus not undergoing hemodynamic testing. It was further shown that ring sizes ≥ 36 mm did not allow for a proper fit of even the largest THV into the ring of all the manufacturers and were consequently not hemodynamically evaluated. The 23 mm THV was too small for any ring size. The lowest gradients were achieved with the 26 mm THV in 30 and 32 mm and the 29 mm THV in 32 and 34 mm rings. CONCLUSION: Not all currently available annuloplasty rings are ideal scaffolds for THV placement. It appears that a more proper fit can be achieved with semi-rigid rings than with rigid ones. Note that 23 mm THV appeared to be too small for an adequate anchoring in even the smallest available ring. Thus, 26 mm as well as 29 mm THV fit properly in ring sizes between 28 and 34 mm. Surgeons may consider to choose from those ring brands and sizes which allow for good placement of a THV in view of possible valve degeneration in the later course.

No Signal Intensity Increase in the Dentate Nucleus on Unenhanced T1-weighted MR Images after More than 20 Serial Injections of Macrocyclic Gadolinium-based Contrast Agents.

Purpose To determine the effect of more than 20 serial injections of macrocyclic gadolinium-based contrast agents (GBCAs) on the signal intensity (SI) of the dentate nucleus (DN) on unenhanced T1-weighted magnetic resonance (MR) images. Materials and Methods In this retrospective, institutional review board-approved study, 33 patients who underwent at least 20 consecutive MR imaging examinations (plus an additional MR imaging for reference) with the exclusive use of macrocyclic GBCAs gadoterate meglumine and gadobutrol were analyzed. SI ratio differences were calculated for DN-to-pons and DN-to-middle cerebellar peduncle (MCP) ratios by subtracting the SI ratio at the first MR imaging examination from the SI ratio at the last MR imaging examination. One-sample t tests were used to examine if the SI ratio differences differed from 0, and Bayes factors were calculated to quantify the strength of evidence for each test. Results Patients underwent a mean of 23.03 ± (standard deviation) 4.20 GBCA administrations (mean accumulated dose, 491.21 mL ± 87.04 of a 0.5 M GBCA solution) with an average of 12.09 weeks ± 2.16 between every administration. Both ratio differences did not differ significantly from 0 (DN-to-pons ratio: -0.0032 ± 0.0154, P = .248; DN-to-MCP ratio: -0.0011 ± 0.0093, P = .521), and one-sided Bayes factors provided substantial to strong evidence against an SI ratio increase (Bayes factor for DN-to-pons ratio = 0.09 and that for DN-to-MCP ratio = 0.12). Conclusion The study indicates that 20 or more serial injections of macrocyclic GBCAs administered with on average 3 months between each injection are not associated with an SI increase in the DN. © RSNA, 2016.

Pediatric Brain: No Increased Signal Intensity in the Dentate Nucleus on Unenhanced T1-weighted MR Images after Consecutive Exposure to a Macrocyclic Gadolinium-based Contrast Agent.

Purpose To determine the effect of at least five serial injections of the macrocyclic gadolinium-based contrast agent (GBCA) gadoterate meglumine on the signal intensity (SI) of the dentate nucleus (DN) of the pediatric brain on nonenhanced T1-weighted magnetic resonance (MR) images. Materials and Methods In this retrospective, institutional review board-approved study, 41 pediatric patients (age range, 3-17 years) who were imaged in at least five consecutive 1.5-T MR examinations with the exclusive use of gadoterate meglumine (plus a final additional nonenhanced MR imaging examination) were evaluated. SI ratio differences between the first and last MR examination were calculated for DN-to-pons and DN-to-middle cerebellar peduncle (MCP) ratios in a region-of-interest-based analysis, and one-sample t tests were used to examine if the SI ratio differences differed from 0. Bayes factors were calculated to quantify the strength of evidence for each test. Results Patients underwent a mean of 8.6 ± 3.9 GBCA administrations (mean accumulated dose, 32.07 mmol ± 17.62, with an average of 16.7 weeks ± 7.9 between every administration). Both ratio differences did not differ significantly from 0 (DN-to-pons ratio: -0.0012 ± 0.0101, P = .436; DN-to-MCP ratio: 0.0007 ± 0.0088, P = .604), and one-sided Bayes factors provided substantial evidence against an SI ratio increase (0.10 for DN-to-pons ratio; 0.27 for DN-to-MCP ratio). Conclusion No increase of the SI in the DN was found after a mean of 8.6 serial injections of the macrocyclic GBCA gadoterate meglumine in pediatric patients, confirming previous studies that did not find this effect after serial injections of macrocyclic GBCAs in adults. © RSNA, 2017.

A Synthetic Route to ß-Hydroxytyrosine-Derived Tetramic Acids: Total Synthesis of the Fungal Metabolite F-14329.

3-Acyltetramic acids derived from ß-hydroxytyrosine are synthetically challenging. The first route to this structural motif, based upon a condensation between a Meldrum's acid conjugate bearing the acyl side chain, and a ß-hydroxytyrosinate, N-protected by an ortho-nitrobenzyl group is presented. This group enables the Dieckmann cyclization of the resulting N-(ß-ketoacyl)amino ester, after which it can be removed photolytically without compromising the delicate 3'-hydroxy group. This strategy was applied to the first total synthesis of the fungal metabolite F-14329 (1).

Computed tomography versus invasive coronary angiography: design and methods of the pragmatic randomised multicentre DISCHARGE trial.

OBJECTIVES: More than 3.5 million invasive coronary angiographies (ICA) are performed in Europe annually. Approximately 2 million of these invasive procedures might be reduced by noninvasive tests because no coronary intervention is performed. Computed tomography (CT) is the most accurate noninvasive test for detection and exclusion of coronary artery disease (CAD). To investigate the comparative effectiveness of CT and ICA, we designed the European pragmatic multicentre DISCHARGE trial funded by the 7th Framework Programme of the European Union (EC-GA 603266). METHODS: In this trial, patients with a low-to-intermediate pretest probability (10-60 %) of suspected CAD and a clinical indication for ICA because of stable chest pain will be randomised in a 1-to-1 ratio to CT or ICA. CT and ICA findings guide subsequent management decisions by the local heart teams according to current evidence and European guidelines. RESULTS: Major adverse cardiovascular events (MACE) defined as cardiovascular death, myocardial infarction and stroke as a composite endpoint will be the primary outcome measure. Secondary and other outcomes include cost-effectiveness, radiation exposure, health-related quality of life (HRQoL), socioeconomic status, lifestyle, adverse events related to CT/ICA, and gender differences. CONCLUSIONS: The DISCHARGE trial will assess the comparative effectiveness of CT and ICA. KEY POINTS: • Coronary artery disease (CAD) is a major cause of morbidity and mortality. • Invasive coronary angiography (ICA) is the reference standard for detection of CAD. • Noninvasive computed tomography angiography excludes CAD with high sensitivity. • CT may effectively reduce the approximately 2 million negative ICAs in Europe. • DISCHARGE addresses this hypothesis in patients with low-to-intermediate pretest probability for CAD.

On the Reliability of Automatic Volume Delineation in Low-Contrast [(18)F]FMISO-PET Imaging.

Hypoxia is a marker of poor prognosis in malignant tumors independent from the selected therapeutic method and the therapy should be intensified in such tumors. Hypoxia imaging with positron emission tomography (PET) is limited by low contrast to noise ratios with every available tracer. In radiation oncology appropriate delineation is required to allow therapy and intensification. While manual segmentation results are highly dependent from experience and observers condition (high inter- and intra observer variability), threshold- and gradient-based algorithms for automatic segmentation frequently fail in low contrast data sets. Likewise, calibration of these algorithms using phantoms is not useful. Complex computational models such as swarm intelligence-based algorithms are promising tools for optimized segmentation results and allow observer independent interpretation of multimodal and multidimensional imaging data.

Spatial distribution of FMISO in head and neck squamous cell carcinomas during radio-chemotherapy and its correlation to pattern of failure.

BACKGROUND: Tumour hypoxia can be measured by FMISO-PET and negatively impacts local tumour control in patients with head and neck squamous cell carcinoma (HNSCC) undergoing radiotherapy. The aim of this post hoc analysis of a prospective clinical trial was to investigate the spatial variability of FMISO hypoxic subvolumes during radio-chemotherapy and the co-localisation of these volumes with later recurrences as a basis for individualised dose prescription trials with dose escalation defined by FMISO-PET. METHODS: Sequential FMISO scans of 12 (of 25) patients presenting residual hypoxia taken before (FMISOpre) and during (FMISOw1-FMISOw5) radio-chemotherapy were analysed regarding the stability of the FMISO subvolumes and, in case of local failure, their correlation to local relapse. RESULTS: Consecutive FMISO-PET positive volumes could be classified as moderately stable with Dice conformity indices of 62% and 58% up to the second week of treatment. Substantial volumetric variation during treatment was observed, with more than 20% geographic miss in all patients and more than 40% in half of the patients. The localisation of the maximum standardised uptake value (SUVmax) differed with a mean distance of 7.0 mm and 13.5 mm between the pre-therapeutic and first or second FMISO-PET during treatment. A stable hypoxic consensual volume (i.e. overlap of pre-therapeutic FMISO and intra-treatment FMISO subvolumes up to week two, generated by different contouring methods) was determined for six patients with imaging information of local recurrence. Three of these six local recurrences were located within this consensual volume. CONCLUSIONS: Our data suggest that selective dose painting to hypoxic tumour subvolumes requires adaptation during treatment and sufficient margins. An alternative strategy is to escalate the dose to the gross tumour volume, accepting lesser escalation of dose outside hypoxic areas if indicated by constraints for organs at risk.

NTCP reduction for advanced head and neck cancer patients using proton therapy for complete or sequential boost treatment versus photon therapy.

BACKGROUND: To determine by treatment plan comparison differences in toxicity risk reduction for patients with head and neck squamous cell carcinoma (HNSCC) from proton therapy either used for complete treatment or sequential boost treatment only. MATERIALS AND METHODS: For 45 HNSCC patients, intensity-modulated photon (IMXT) and proton (IMPT) treatment plans were created including a dose escalation via simultaneous integrated boost with a one-step adaptation strategy after 25 fractions for sequential boost treatment. Dose accumulation was performed for pure IMXT treatment, pure IMPT treatment and for a mixed modality treatment with IMXT for the elective target followed by a sequential boost with IMPT. Treatment plan evaluation was based on modern normal tissue complication probability (NTCP) models for mucositis, xerostomia, aspiration, dysphagia, larynx edema and trismus. Individual NTCP differences between IMXT and IMPT (∆NTCPIMXT-IMPT) as well as between IMXT and the mixed modality treatment (∆NTCPIMXT-Mix) were calculated. RESULTS: Target coverage was similar in all three scenarios. NTCP values could be reduced in all patients using IMPT treatment. However, ∆NTCPIMXT-Mix values were a factor 2-10 smaller than ∆NTCPIMXT-IMPT. Assuming a threshold of ≥ 10% NTCP reduction in xerostomia or dysphagia risk as criterion for patient assignment to IMPT, less than 15% of the patients would be selected for a proton boost, while about 50% would be assigned to pure IMPT treatment. For mucositis and trismus, ∆NTCP ≥ 10% occurred in six and four patients, respectively, with pure IMPT treatment, while no such difference was identified with the proton boost. CONCLUSIONS: The use of IMPT generally reduces the expected toxicity risk while maintaining good tumor coverage in the examined HNSCC patients. A mixed modality treatment using IMPT solely for a sequential boost reduces the risk by 10% only in rare cases. In contrast, pure IMPT treatment may be reasonable for about half of the examined patient cohort considering the toxicities xerostomia and dysphagia, if a feasible strategy for patient anatomy changes is implemented.

A neurosurgical phantom-based training system with ultrasound simulation.

BACKGROUND: Brain tumor surgeries are associated with a high technical and personal effort. The required interactions between the surgeon and the technical components, such as neuronavigation, surgical instruments and intraoperative imaging, are complex and demand innovative training solutions and standardized evaluation methods. Phantom-based training systems could be useful in complementing the existing surgical education and training. METHODS: A prototype of a phantom-based training system was developed, intended for standardized training of important aspects of brain tumor surgery based on real patient data. The head phantom consists of a three-part construction that includes a reusable base and adapter, as well as a changeable module for single use. Training covers surgical planning of the optimal access path, the setup of the navigation system including the registration of the head phantom, as well as the navigated craniotomy with real instruments. Tracked instruments during the simulation and predefined access paths constitute the basis for the essential objective training feedback. RESULTS: The prototype was evaluated in a pilot study by assistant physicians at different education levels. They performed a complete simulation and a final assessment using an evaluation questionnaire. The analysis of the questionnaire showed the evaluation result as "good" for the phantom construction and the used materials. The learning effect concerning the navigated planning was evaluated as "very good", as well as having the effect of increasing safety for the surgeon before planning and conducting craniotomies independently on patients. CONCLUSIONS: The training system represents a promising approach for the future training of neurosurgeons. It aims to improve surgical skill training by creating a more realistic simulation in a non-risk environment. Hence, it could help to bridge the gap between theoretical and practical training with the potential to benefit both physicians and patients.

Automated quantification of photoreceptor outer segments in developing and degenerating retinas on microscopy images across scales.

The functionality of photoreceptors, rods, and cones is highly dependent on their outer segments (POS), a cellular compartment containing highly organized membranous structures that generate biochemical signals from incident light. While POS formation and degeneration are qualitatively assessed on microscopy images, reliable methodology for quantitative analyses is still limited. Here, we developed methods to quantify POS (QuaPOS) maturation and quality on retinal sections using automated image analyses. POS formation was examined during the development and in adulthood of wild-type mice via light microscopy (LM) and transmission electron microscopy (TEM). To quantify the number, size, shape, and fluorescence intensity of POS, retinal cryosections were immunostained for the cone POS marker S-opsin. Fluorescence images were used to train the robust classifier QuaPOS-LM based on supervised machine learning for automated image segmentation. Characteristic features of segmentation results were extracted to quantify the maturation of cone POS. Subsequently, this quantification method was applied to characterize POS degeneration in "cone photoreceptor function loss 1" mice. TEM images were used to establish the ultrastructural quantification method QuaPOS-TEM for the alignment of POS membranes. Images were analyzed using a custom-written MATLAB code to extract the orientation of membranes from the image gradient and their alignment (coherency). This analysis was used to quantify the POS morphology of wild-type and two inherited retinal degeneration ("retinal degeneration 19" and "rhodopsin knock-out") mouse lines. Both automated analysis technologies provided robust characterization and quantification of POS based on LM or TEM images. Automated image segmentation by the classifier QuaPOS-LM and analysis of the orientation of membrane stacks by QuaPOS-TEM using fluorescent or TEM images allowed quantitative evaluation of POS formation and quality. The assessments showed an increase in POS number, volume, and membrane coherency during wild-type postnatal development, while a decrease in all three observables was detected in different retinal degeneration mouse models. All the code used for the presented analysis is open source, including example datasets to reproduce the findings. Hence, the QuaPOS quantification methods are useful for in-depth characterization of POS on retinal sections in developmental studies, for disease modeling, or after therapeutic interventions affecting photoreceptors.