On the microstructural origin of brain white matter hydraulic permeability.

Brain microstructure plays a key role in driving the transport of drug molecules directly administered to the brain tissue, as in Convection-Enhanced Delivery procedures. The proposed research analyzes the hydraulic permeability of two white matter (WM) areas (corpus callosum and fornix) whose three-dimensional microstructure was reconstructed starting from the acquisition of electron microscopy images. We cut the two volumes with 20 equally spaced planes distributed along two perpendicular directions, and, on each plane, we computed the corresponding permeability vector. Then, we considered that the WM structure is mainly composed of elongated and parallel axons, and, using a principal component analysis, we defined two principal directions, parallel and perpendicular, with respect to the axons' main direction. The latter were used to define a reference frame onto which the permeability vectors were projected to finally obtain the permeability along the parallel and perpendicular directions. The results show a statistically significant difference between parallel and perpendicular permeability, with a ratio of about two in both the WM structures analyzed, thus demonstrating their anisotropic behavior. Moreover, we find a significant difference between permeability in corpus callosum and fornix, which suggests that the WM heterogeneity should also be considered when modeling drug transport in the brain. Our findings, which demonstrate and quantify the anisotropic and heterogeneous character of the WM, represent a fundamental contribution not only for drug-delivery modeling, but also for shedding light on the interstitial transport mechanisms in the extracellular space.

EMG-driven control in lower limb prostheses: a topic-based systematic review.

BACKGROUND: The inability of users to directly and intuitively control their state-of-the-art commercial prosthesis contributes to a low device acceptance rate. Since Electromyography (EMG)-based control has the potential to address those inabilities, research has flourished on investigating its incorporation in microprocessor-controlled lower limb prostheses (MLLPs). However, despite the proposed benefits of doing so, there is no clear explanation regarding the absence of a commercial product, in contrast to their upper limb counterparts. OBJECTIVE AND METHODOLOGIES: This manuscript aims to provide a comparative overview of EMG-driven control methods for MLLPs, to identify their prospects and limitations, and to formulate suggestions on future research and development. This is done by systematically reviewing academical studies on EMG MLLPs. In particular, this review is structured by considering four major topics: (1) type of neuro-control, which discusses methods that allow the nervous system to control prosthetic devices through the muscles; (2) type of EMG-driven controllers, which defines the different classes of EMG controllers proposed in the literature; (3) type of neural input and processing, which describes how EMG-driven controllers are implemented; (4) type of performance assessment, which reports the performance of the current state of the art controllers. RESULTS AND CONCLUSIONS: The obtained results show that the lack of quantitative and standardized measures hinders the possibility to analytically compare the performances of different EMG-driven controllers. In relation to this issue, the real efficacy of EMG-driven controllers for MLLPs have yet to be validated. Nevertheless, in anticipation of the development of a standardized approach for validating EMG MLLPs, the literature suggests that combining multiple neuro-controller types has the potential to develop a more seamless and reliable EMG-driven control. This solution has the promise to retain the high performance of the currently employed non-EMG-driven controllers for rhythmic activities such as walking, whilst improving the performance of volitional activities such as task switching or non-repetitive movements. Although EMG-driven controllers suffer from many drawbacks, such as high sensitivity to noise, recent progress in invasive neural interfaces for prosthetic control (bionics) will allow to build a more reliable connection between the user and the MLLPs. Therefore, advancements in powered MLLPs with integrated EMG-driven control have the potential to strongly reduce the effects of psychosomatic conditions and musculoskeletal degenerative pathologies that are currently affecting lower limb amputees.

Knowledge-based automated planning system for StereoElectroEncephaloGraphy: A center-based scenario.

Surgical planning for StereoElectroEncephaloGraphy (SEEG) is a complex and patient specific task, where the experience and medical workflow of each institution may influence the final planning choices. To account for this variability, we developed a data-based Computer Assisted Planning (CAP) solution able to exploit the knowledge extracted by past cases. By the analysis of retrospective patients' data sets, our system proposes a pool of trajectories commonly used by the institution, which can be selected to initialize a new patient plan. An optimization framework adapts those to the patient's anatomy by optimizing clinical requirements (e.g. distance from vessel, gray matter recording and insertion angle), and adapting its strategy based on the trajectory type selected.The system has been customized based on the data of a single institution. Two neurosurgeons, working in a high-volume hospital, have validated it by using 15 retrospective patient data sets, with more than 200 trajectories reviewed. Both surgeons considered ~81% of the optimized trajectories as clinically feasible (75% inter-rater reliability). Quantitative comparison of distance from vessels, insertion angle and gray matter recording index showed that the optimized trajectories reached superior or comparable values with respect to the original manual plans. The results suggest that a tailored center-based solution could increase the acceptance rate of the automated trajectories proposed.

An Online Method to Detect and Locate an External Load on the Human Body with Applications in Ergonomics Assessment.

In this work, we propose an online method to detect and approximately locate an external load induced on the body of a person interacting with the environment. The method is based on a torque equilibrium condition on the human sagittal plane, which takes into account a reduced-complexity model of the whole-body centre of pressure (CoP) along with the measured one, and the vertical component of the ground reaction forces (vGRFs). The latter is combined with a statistical analysis approach to improve the localisation accuracy, (which is subject to uncertainties) to the extent of the industrial applications we target. The proposed technique eliminates the assumption of known contact position of an external load on the human limbs, allowing a more flexible online body-state tracking. The accuracy of the proposed method is first evaluated via a simulation study in which various contact points on different body postures are considered. Next, experiments on human subjects with three different contact locations applied to the human body are presented, revealing the validity of the proposed methodology. Lastly, its benefit in the estimation of human dynamic states is demonstrated. These results add another layer to the online human ergonomics assessment framework developed in our laboratory, extending it to more realistic and varying interaction conditions.

A Fast and Robust Deep Convolutional Neural Networks for Complex Human Activity Recognition Using Smartphone.

As a significant role in healthcare and sports applications, human activity recognition (HAR) techniques are capable of monitoring humans' daily behavior. It has spurred the demand for intelligent sensors and has been giving rise to the explosive growth of wearable and mobile devices. They provide the most availability of human activity data (big data). Powerful algorithms are required to analyze these heterogeneous and high-dimension streaming data efficiently. This paper proposes a novel fast and robust deep convolutional neural network structure (FR-DCNN) for human activity recognition (HAR) using a smartphone. It enhances the effectiveness and extends the information of the collected raw data from the inertial measurement unit (IMU) sensors by integrating a series of signal processing algorithms and a signal selection module. It enables a fast computational method for building the DCNN classifier by adding a data compression module. Experimental results on the sampled 12 complex activities dataset show that the proposed FR-DCNN model is the best method for fast computation and high accuracy recognition. The FR-DCNN model only needs 0.0029 s to predict activity in an online way with 95.27% accuracy. Meanwhile, it only takes 88 s (average) to establish the DCNN classifier on the compressed dataset with less precision loss 94.18%.

Towards Model-Free Tool Dynamic Identification and Calibration Using Multi-Layer Neural Network.

In robot control with physical interaction, like robot-assisted surgery and bilateral teleoperation, the availability of reliable interaction force information has proved to be capable of increasing the control precision and of dealing with the surrounding complex environments. Usually, force sensors are mounted between the end effector of the robot manipulator and the tool for measuring the interaction forces on the tooltip. In this case, the force acquired from the force sensor includes not only the interaction force but also the gravity force of the tool. Hence the tool dynamic identification is required for accurate dynamic simulation and model-based control. Although model-based techniques have already been widely used in traditional robotic arms control, their accuracy is limited due to the lack of specific dynamic models. This work proposes a model-free technique for dynamic identification using multi-layer neural networks (MNN). It utilizes two types of MNN architectures based on both feed-forward networks (FF-MNN) and cascade-forward networks (CF-MNN) to model the tool dynamics. Compared with the model-based technique, i.e., curve fitting (CF), the accuracy of the tool identification is improved. After the identification and calibration, a further demonstration of bilateral teleoperation is presented using a serial robot (LWR4+, KUKA, Germany) and a haptic manipulator (SIGMA 7, Force Dimension, Switzerland). Results demonstrate the promising performance of the model-free tool identification technique using MNN, improving the results provided by model-based methods.

Enhanced Real-Time Pose Estimation for Closed-Loop Robotic Manipulation of Magnetically Actuated Capsule Endoscopes.

Pose estimation methods for robotically guided magnetic actuation of capsule endoscopes have recently enabled trajectory following and automation of repetitive endoscopic maneuvers. However, these methods face significant challenges in their path to clinical adoption including the presence of regions of magnetic field singularity, where the accuracy of the system degrades, and the need for accurate initialization of the capsule's pose. In particular, the singularity problem exists for any pose estimation method that utilizes a single source of magnetic field if the method does not rely on the motion of the magnet to obtain multiple measurements from different vantage points. We analyze the workspace of such pose estimation methods with the use of the point-dipole magnetic field model and show that singular regions exist in areas where the capsule is nominally located during magnetic actuation. Since the dipole model can approximate most magnetic field sources, the problem discussed herein pertains to a wider set of pose estimation techniques. We then propose a novel hybrid approach employing static and time-varying magnetic field sources and show that this system has no regions of singularity. The proposed system was experimentally validated for accuracy, workspace size, update rate and performance in regions of magnetic singularity. The system performed as well or better than prior pose estimation methods without requiring accurate initialization and was robust to magnetic singularity. Experimental demonstration of closed-loop control of a tethered magnetic device utilizing the developed pose estimation technique is provided to ascertain its suitability for robotically guided capsule endoscopy. Hence, advances in closed-loop control and intelligent automation of magnetically actuated capsule endoscopes can be further pursued toward clinical realization by employing this pose estimation system.

A new tool for touch-free patient registration for robot-assisted intracranial surgery: application accuracy from a phantom study and a retrospective surgical series.

OBJECTIVE The purpose of this study was to compare the accuracy of Neurolocate frameless registration system and frame-based registration for robotic stereoelectroencephalography (SEEG). METHODS The authors performed a 40-trajectory phantom laboratory study and a 127-trajectory retrospective analysis of a surgical series. The laboratory study was aimed at testing the noninferiority of the Neurolocate system. The analysis of the surgical series compared Neurolocate-based SEEG implantations with a frame-based historical control group. RESULTS The mean localization errors (LE) ± standard deviations (SD) for Neurolocate-based and frame-based trajectories were 0.67 ± 0.29 mm and 0.76 ± 0.34 mm, respectively, in the phantom study (p = 0.35). The median entry point LE was 0.59 mm (interquartile range [IQR] 0.25-0.88 mm) for Neurolocate-registration-based trajectories and 0.78 mm (IQR 0.49-1.08 mm) for frame-registration-based trajectories (p = 0.00002) in the clinical study. The median target point LE was 1.49 mm (IQR 1.06-2.4 mm) for Neurolocate-registration-based trajectories and 1.77 mm (IQR 1.25-2.5 mm) for frame-registration-based trajectories in the clinical study. All the surgical procedures were successful and uneventful. CONCLUSIONS The results of the phantom study demonstrate the noninferiority of Neurolocate frameless registration. The results of the retrospective surgical series analysis suggest that Neurolocate-based procedures can be more accurate than the frame-based ones. The safety profile of Neurolocate-based registration should be similar to that of frame-based registration. The Neurolocate system is comfortable, noninvasive, easy to use, and potentially faster than other registration devices.

Learning-based keypoint registration for fetoscopic mosaicking.

PURPOSE: In twin-to-twin transfusion syndrome (TTTS), abnormal vascular anastomoses in the monochorionic placenta can produce uneven blood flow between the two fetuses. In the current practice, TTTS is treated surgically by closing abnormal anastomoses using laser ablation. This surgery is minimally invasive and relies on fetoscopy. Limited field of view makes anastomosis identification a challenging task for the surgeon. METHODS: To tackle this challenge, we propose a learning-based framework for in vivo fetoscopy frame registration for field-of-view expansion. The novelties of this framework rely on a learning-based keypoint proposal network and an encoding strategy to filter (i) irrelevant keypoints based on fetoscopic semantic image segmentation and (ii) inconsistent homographies. RESULTS: We validate our framework on a dataset of six intraoperative sequences from six TTTS surgeries from six different women against the most recent state-of-the-art algorithm, which relies on the segmentation of placenta vessels. CONCLUSION: The proposed framework achieves higher performance compared to the state of the art, paving the way for robust mosaicking to provide surgeons with context awareness during TTTS surgery.

Spatio-temporal layers based intra-operative stereo depth estimation network via hierarchical prediction and progressive training.

BACKGROUND AND OBJECTIVE: Safety of robotic surgery can be enhanced through augmented vision or artificial constraints to the robotl motion, and intra-operative depth estimation is the cornerstone of these applications because it provides precise position information of surgical scenes in 3D space. High-quality depth estimation of endoscopic scenes has been a valuable issue, and the development of deep learning provides more possibility and potential to address this issue. METHODS: In this paper, a deep learning-based approach is proposed to recover 3D information of intra-operative scenes. To this aim, a fully 3D encoder-decoder network integrating spatio-temporal layers is designed, and it adopts hierarchical prediction and progressive learning to enhance prediction accuracy and shorten training time. RESULTS: Our network gets the depth estimation accuracy of MAE 2.55±1.51 (mm) and RMSE 5.23±1.40 (mm) using 8 surgical videos with a resolution of 1280×1024, which performs better compared with six other state-of-the-art methods that were trained on the same data. CONCLUSIONS: Our network can implement a promising depth estimation performance in intra-operative scenes using stereo images, allowing the integration in robot-assisted surgery to enhance safety.

Towards safer robot-assisted surgery: A markerless augmented reality framework.

Robot-assisted surgery is rapidly developing in the medical field, and the integration of augmented reality shows the potential to improve the operation performance of surgeons by providing more visual information. In this paper, we proposed a markerless augmented reality framework to enhance safety by avoiding intra-operative bleeding, which is a high risk caused by collision between surgical instruments and delicate blood vessels (arteries or veins). Advanced stereo reconstruction and segmentation networks are compared to find the best combination to reconstruct the intra-operative blood vessel in 3D space for registration with the pre-operative model, and the minimum distance detection between the instruments and the blood vessel is implemented. A robot-assisted lymphadenectomy is emulated on the da Vinci Research Kit in a dry lab, and ten human subjects perform this operation to explore the usability of the proposed framework. The result shows that the augmented reality framework can help the users to avoid the dangerous collision between the instruments and the delicate blood vessel while not introducing an extra load. It provides a flexible framework that integrates augmented reality into the medical robotic platform to enhance safety during surgery.

Path tracking control of a steerable catheter in transcatheter cardiology interventions.

PURPOSE: Intracardiac transcatheter interventions allow for reducing trauma and hospitalization stays as compared to standard surgery. In the treatment of mitral regurgitation, the most widely adopted transcatheter approach consists in deploying a clip on the mitral valve leaflets by means of a catheter that is run through veins from a peripheral access to the left atrium. However, precise manipulation of the catheter from outside the body while copying with the path constraints imposed by the vessels remains challenging. METHODS: We proposed a path tracking control framework that provides adequate motion commands to the robotic steerable catheter for autonomous navigation through vascular lumens. The proposed work implements a catheter kinematic model featuring nonholonomic constraints. Relying on the real-time measurements from an electromagnetic sensor and a fiber Bragg grating sensor, a two-level feedback controller was designed to control the catheter. RESULTS: The proposed method was tested in a patient-specific vessel phantom. A median position error between the center line of the vessel and the catheter tip trajectory was found to be below 2 mm, with a maximum error below 3 mm. Statistical testing confirmed that the performance of the proposed method exhibited no significant difference in both free space and the contact region. CONCLUSION: The preliminary in vitro studies presented in this paper showed promising accuracy in navigating the catheter within the vessel. The proposed approach enables autonomous control of a steerable catheter for transcatheter cardiology interventions without the request of calibrating the intuitive parameters or acquiring a training dataset.

Robotic systems for upper-limb rehabilitation in multiple sclerosis: a SWOT analysis and the synergies with virtual and augmented environments.

The robotics discipline is exploring precise and versatile solutions for upper-limb rehabilitation in Multiple Sclerosis (MS). People with MS can greatly benefit from robotic systems to help combat the complexities of this disease, which can impair the ability to perform activities of daily living (ADLs). In order to present the potential and the limitations of smart mechatronic devices in the mentioned clinical domain, this review is structured to propose a concise SWOT (Strengths, Weaknesses, Opportunities, and Threats) Analysis of robotic rehabilitation in MS. Through the SWOT Analysis, a method mostly adopted in business management, this paper addresses both internal and external factors that can promote or hinder the adoption of upper-limb rehabilitation robots in MS. Subsequently, it discusses how the synergy with another category of interaction technologies - the systems underlying virtual and augmented environments - may empower Strengths, overcome Weaknesses, expand Opportunities, and handle Threats in rehabilitation robotics for MS. The impactful adaptability of these digital settings (extensively used in rehabilitation for MS, even to approach ADL-like tasks in safe simulated contexts) is the main reason for presenting this approach to face the critical issues of the aforementioned SWOT Analysis. This methodological proposal aims at paving the way for devising further synergistic strategies based on the integration of medical robotic devices with other promising technologies to help upper-limb functional recovery in MS.

Placental vessel segmentation and registration in fetoscopy: Literature review and MICCAI FetReg2021 challenge findings.

Fetoscopy laser photocoagulation is a widely adopted procedure for treating Twin-to-Twin Transfusion Syndrome (TTTS). The procedure involves photocoagulation pathological anastomoses to restore a physiological blood exchange among twins. The procedure is particularly challenging, from the surgeon's side, due to the limited field of view, poor manoeuvrability of the fetoscope, poor visibility due to amniotic fluid turbidity, and variability in illumination. These challenges may lead to increased surgery time and incomplete ablation of pathological anastomoses, resulting in persistent TTTS. Computer-assisted intervention (CAI) can provide TTTS surgeons with decision support and context awareness by identifying key structures in the scene and expanding the fetoscopic field of view through video mosaicking. Research in this domain has been hampered by the lack of high-quality data to design, develop and test CAI algorithms. Through the Fetoscopic Placental Vessel Segmentation and Registration (FetReg2021) challenge, which was organized as part of the MICCAI2021 Endoscopic Vision (EndoVis) challenge, we released the first large-scale multi-center TTTS dataset for the development of generalized and robust semantic segmentation and video mosaicking algorithms with a focus on creating drift-free mosaics from long duration fetoscopy videos. For this challenge, we released a dataset of 2060 images, pixel-annotated for vessels, tool, fetus and background classes, from 18 in-vivo TTTS fetoscopy procedures and 18 short video clips of an average length of 411 frames for developing placental scene segmentation and frame registration for mosaicking techniques. Seven teams participated in this challenge and their model performance was assessed on an unseen test dataset of 658 pixel-annotated images from 6 fetoscopic procedures and 6 short clips. For the segmentation task, overall baseline performed was the top performing (aggregated mIoU of 0.6763) and was the best on the vessel class (mIoU of 0.5817) while team RREB was the best on the tool (mIoU of 0.6335) and fetus (mIoU of 0.5178) classes. For the registration task, overall the baseline performed better than team SANO with an overall mean 5-frame SSIM of 0.9348. Qualitatively, it was observed that team SANO performed better in planar scenarios, while baseline was better in non-planner scenarios. The detailed analysis showed that no single team outperformed on all 6 test fetoscopic videos. The challenge provided an opportunity to create generalized solutions for fetoscopic scene understanding and mosaicking. In this paper, we present the findings of the FetReg2021 challenge, alongside reporting a detailed literature review for CAI in TTTS fetoscopy. Through this challenge, its analysis and the release of multi-center fetoscopic data, we provide a benchmark for future research in this field.

FUN-SIS: A Fully UNsupervised approach for Surgical Instrument Segmentation.

Automatic surgical instrument segmentation of endoscopic images is a crucial building block of many computer-assistance applications for minimally invasive surgery. So far, state-of-the-art approaches completely rely on the availability of a ground-truth supervision signal, obtained via manual annotation, thus expensive to collect at large scale. In this paper, we present FUN-SIS, a Fully-UNsupervised approach for binary Surgical Instrument Segmentation. FUN-SIS trains a per-frame segmentation model on completely unlabelled endoscopic videos, by solely relying on implicit motion information and instrument shape-priors. We define shape-priors as realistic segmentation masks of the instruments, not necessarily coming from the same dataset/domain as the videos. The shape-priors can be collected in various and convenient ways, such as recycling existing annotations from other datasets. We leverage them as part of a novel generative-adversarial approach, allowing to perform unsupervised instrument segmentation of optical-flow images during training. We then use the obtained instrument masks as pseudo-labels in order to train a per-frame segmentation model; to this aim, we develop a learning-from-noisy-labels architecture, designed to extract a clean supervision signal from these pseudo-labels, leveraging their peculiar noise properties. We validate the proposed contributions on three surgical datasets, including the MICCAI 2017 EndoVis Robotic Instrument Segmentation Challenge dataset. The obtained fully-unsupervised results for surgical instrument segmentation are almost on par with the ones of fully-supervised state-of-the-art approaches. This suggests the tremendous potential of the proposed method to leverage the great amount of unlabelled data produced in the context of minimally invasive surgery.

Toward a navigation framework for fetoscopy.

PURPOSE: Fetoscopic laser photocoagulation of placental anastomoses is the most effective treatment for twin-to-twin transfusion syndrome (TTTS). A robust mosaic of placenta and its vascular network could support surgeons' exploration of the placenta by enlarging the fetoscope field-of-view. In this work, we propose a learning-based framework for field-of-view expansion from intra-operative video frames. METHODS: While current state of the art for fetoscopic mosaicking builds upon the registration of anatomical landmarks which may not always be visible, our framework relies on learning-based features and keypoints, as well as robust transformer-based image-feature matching, without requiring any anatomical priors. We further address the problem of occlusion recovery and frame relocalization, relying on the computed features and their descriptors. RESULTS: Experiments were conducted on 10 in-vivo TTTS videos from two different fetal surgery centers. The proposed framework was compared with several state-of-the-art approaches, achieving higher [Formula: see text] on 7 out of 10 videos and a success rate of [Formula: see text] in occlusion recovery. CONCLUSION: This work introduces a learning-based framework for placental mosaicking with occlusion recovery from intra-operative videos using a keypoint-based strategy and features. The proposed framework can compute the placental panorama and recover even in case of camera tracking loss where other methods fail. The results suggest that the proposed framework has large potential to pave the way to creating a surgical navigation system for TTTS by providing robust field-of-view expansion.

Human-Centered Functional Task Design for Robotic Upper-Limb Rehabilitation.

Robotic rehabilitation has demonstrated slight positive effects compared to traditional care, but there is still a lack of targeted high-level control strategies in the current state-of-the-art for minimizing pathological motor behaviors. In this study, we analyzed upper-limb motion capture data from healthy subjects performing a pick-and-place task to identify task-specific variability in postural patterns. The results revealed consistent behaviors among subjects, presenting an opportunity to develop a novel extraction method for variable volume references based solely on observations from healthy individuals. These human-centered references were tested on a simulated 4 degrees-of-freedom upper-limb exoskeleton, showing its compliant adaptation to the path considering the variance in healthy subjects' motor behavior.

FRSR: Framework for real-time scene reconstruction in robot-assisted minimally invasive surgery.

3D reconstruction of the intra-operative scenes provides precise position information which is the foundation of various safety related applications in robot-assisted surgery, such as augmented reality. Herein, a framework integrated into a known surgical system is proposed to enhance the safety of robotic surgery. In this paper, we present a scene reconstruction framework to restore the 3D information of the surgical site in real time. In particular, a lightweight encoder-decoder network is designed to perform disparity estimation, which is the key component of the scene reconstruction framework. The stereo endoscope of da Vinci Research Kit (dVRK) is adopted to explore the feasibility of the proposed approach, and it provides the possibility for the migration to other Robot Operating System (ROS) based robot platforms due to the strong independence on hardware. The framework is evaluated using three different scenarios, including a public dataset (3018 pairs of endoscopic images), the scene from the dVRK endoscope in our lab as well as a self-made clinical dataset captured from an oncology hospital. Experimental results show that the proposed framework can reconstruct 3D surgical scenes in real time (25 FPS), and achieve high accuracy (2.69 ± 1.48 mm in MAE, 5.47 ± 1.34 mm in RMSE and 0.41 ± 0.23 in SRE, respectively). It demonstrates that our framework can reconstruct intra-operative scenes with high reliability of both accuracy and speed, and the validation of clinical data also shows its potential in surgery. This work enhances the state of art in 3D intra-operative scene reconstruction based on medical robot platforms. The clinical dataset has been released to promote the development of scene reconstruction in the medical image community.

Semi-Supervised Bladder Tissue Classification in Multi-Domain Endoscopic Images.

OBJECTIVE: Accurate visual classification of bladder tissue during Trans-Urethral Resection of Bladder Tumor (TURBT) procedures is essential to improve early cancer diagnosis and treatment. During TURBT interventions, White Light Imaging (WLI) and Narrow Band Imaging (NBI) techniques are used for lesion detection. Each imaging technique provides diverse visual information that allows clinicians to identify and classify cancerous lesions. Computer vision methods that use both imaging techniques could improve endoscopic diagnosis. We address the challenge of tissue classification when annotations are available only in one domain, in our case WLI, and the endoscopic images correspond to an unpaired dataset, i.e. there is no exact equivalent for every image in both NBI and WLI domains. METHOD: We propose a semi-surprised Generative Adversarial Network (GAN)-based method composed of three main components: a teacher network trained on the labeled WLI data; a cycle-consistency GAN to perform unpaired image-to-image translation, and a multi-input student network. To ensure the quality of the synthetic images generated by the proposed GAN we perform a detailed quantitative, and qualitative analysis with the help of specialists. CONCLUSION: The overall average classification accuracy, precision, and recall obtained with the proposed method for tissue classification are 0.90, 0.88, and 0.89 respectively, while the same metrics obtained in the unlabeled domain (NBI) are 0.92, 0.64, and 0.94 respectively. The quality of the generated images is reliable enough to deceive specialists. SIGNIFICANCE: This study shows the potential of using semi-supervised GAN-based bladder tissue classification when annotations are limited in multi-domain data.

Radiomics-Based Machine Learning Model for Predicting Overall and Progression-Free Survival in Rare Cancer: A Case Study for Primary CNS Lymphoma Patients.

Primary Central Nervous System Lymphoma (PCNSL) is an aggressive neoplasm with a poor prognosis. Although therapeutic progresses have significantly improved Overall Survival (OS), a number of patients do not respond to HD-MTX-based chemotherapy (15-25%) or experience relapse (25-50%) after an initial response. The reasons underlying this poor response to therapy are unknown. Thus, there is an urgent need to develop improved predictive models for PCNSL. In this study, we investigated whether radiomics features can improve outcome prediction in patients with PCNSL. A total of 80 patients diagnosed with PCNSL were enrolled. A patient sub-group, with complete Magnetic Resonance Imaging (MRI) series, were selected for the stratification analysis. Following radiomics feature extraction and selection, different Machine Learning (ML) models were tested for OS and Progression-free Survival (PFS) prediction. To assess the stability of the selected features, images from 23 patients scanned at three different time points were used to compute the Interclass Correlation Coefficient (ICC) and to evaluate the reproducibility of each feature for both original and normalized images. Features extracted from Z-score normalized images were significantly more stable than those extracted from non-normalized images with an improvement of about 38% on average (p-value < 10-12). The area under the ROC curve (AUC) showed that radiomics-based prediction overcame prediction based on current clinical prognostic factors with an improvement of 23% for OS and 50% for PFS, respectively. These results indicate that radiomics features extracted from normalized MR images can improve prognosis stratification of PCNSL patients and pave the way for further study on its potential role to drive treatment choice.