Haploinsufficiency of phosphodiesterase 10A activates PI3K/AKT signaling independent of PTEN to induce an aggressive glioma phenotype.

Glioblastoma is universally fatal and characterized by frequent chromosomal copy number alterations harboring oncogenes and tumor suppressors. In this study, we analyzed exome-wide human glioblastoma copy number data and found that cytoband 6q27 is an independent poor prognostic marker in multiple data sets. We then combined CRISPR-Cas9 data, human spatial transcriptomic data, and human and mouse RNA sequencing data to nominate PDE10A as a potential haploinsufficient tumor suppressor in the 6q27 region. Mouse glioblastoma modeling using the RCAS/tv-a system confirmed that Pde10a suppression induced an aggressive glioma phenotype in vivo and resistance to temozolomide and radiation therapy in vitro. Cell culture analysis showed that decreased Pde10a expression led to increased PI3K/AKT signaling in a Pten-independent manner, a response blocked by selective PI3K inhibitors. Single-nucleus RNA sequencing from our mouse gliomas in vivo, in combination with cell culture validation, further showed that Pde10a suppression was associated with a proneural-to-mesenchymal transition that exhibited increased cell adhesion and decreased cell migration. Our results indicate that glioblastoma patients harboring PDE10A loss have worse outcomes and potentially increased sensitivity to PI3K inhibition.

The Implant-Induced Foreign Body Response Is Limited by CD13-Dependent Regulation of Ubiquitination of Fusogenic Proteins.

Implanted medical devices, from artificial heart valves and arthroscopic joints to implantable sensors, often induce a foreign body response (FBR), a form of chronic inflammation resulting from the inflammatory reaction to a persistent foreign stimulus. The FBR is characterized by a subset of multinucleated giant cells (MGCs) formed by macrophage fusion, the foreign body giant cells (FBGCs), accompanied by inflammatory cytokines, matrix deposition, and eventually deleterious fibrotic implant encapsulation. Despite efforts to improve biocompatibility, implant-induced FBR persists, compromising the utility of devices and making efforts to control the FBR imperative for long-term function. Controlling macrophage fusion in FBGC formation presents a logical target to prevent implant failure, but the actual contribution of FBGCs to FBR-induced damage is controversial. CD13 is a molecular scaffold, and in vitro induction of CD13KO bone marrow progenitors generates many more MGCs than the wild type, suggesting that CD13 regulates macrophage fusion. In the mesh implant model of FBR, CD13KO mice produced significantly more peri-implant FBGCs with enhanced TGF-ß expression and increased collagen deposition versus the wild type. Prior to fusion, increased protrusion and microprotrusion formation accompanies hyperfusion in the absence of CD13. Expression of fusogenic proteins driving cell-cell fusion was aberrantly sustained at high levels in CD13KO MGCs, which we show is due to a novel CD13 function, to our knowledge, regulating ubiquitin/proteasomal protein degradation. We propose CD13 as a physiologic brake limiting aberrant macrophage fusion and the FBR, and it may be a novel therapeutic target to improve the success of implanted medical devices. Furthermore, our data directly implicate FBGCs in the detrimental fibrosis that characterizes the FBR.

CD13 facilitates immune cell migration and aggravates acute injury but promotes chronic post-stroke recovery.

INTRODUCTION: Acute stroke leads to the activation of myeloid cells. These cells express adhesion molecules and transmigrate to the brain, thereby aggravating injury. Chronically after stroke, repair processes, including angiogenesis, are activated and enhance post-stroke recovery. Activated myeloid cells express CD13, which facilitates their migration into the site of injury. However, angiogenic blood vessels which play a role in recovery also express CD13. Overall, the specific contribution of CD13 to acute and chronic stroke outcomes is unknown. METHODS: CD13 expression was estimated in both mice and humans after the ischemic stroke. Young (8-12 weeks) male wild-type and global CD13 knockout (KO) mice were used for this study. Mice underwent 60 min of middle cerebral artery occlusion (MCAO) followed by reperfusion. For acute studies, the mice were euthanized at either 24- or 72 h post-stroke. For chronic studies, the Y-maze, Barnes maze, and the open field were performed on day 7 and day 28 post-stroke. Mice were euthanized at day 30 post-stroke and the brains were collected for assessment of inflammation, white matter injury, tissue loss, and angiogenesis. Flow cytometry was performed on days 3 and 7 post-stroke to quantify infiltrated monocytes and neutrophils and CXCL12/CXCR4 signaling. RESULTS: Brain CD13 expression and infiltrated CD13+ monocytes and neutrophils increased acutely after the stroke. The brain CD13+lectin+ blood vessels increased on day 15 after the stroke. Similarly, an increase in the percentage area CD13 was observed in human stroke patients at the subacute time after stroke. Deletion of CD13 resulted in reduced infarct volume and improved neurological recovery after acute stroke. However, CD13KO mice had significantly worse memory deficits, amplified gliosis, and white matter damage compared to wild-type animals at chronic time points. CD13-deficient mice had an increased percentage of CXCL12+cells but a reduced percentage of CXCR4+cells and decreased angiogenesis at day 30 post-stroke. CONCLUSIONS: CD13 is involved in the trans-migration of monocytes and neutrophils after stroke, and acutely, led to decreased infarct size and improved behavioral outcomes. However, loss of CD13 led to reductions in post-stroke angiogenesis by reducing CXCL12/CXCR4 signaling.

VSGD-Net: Virtual Staining Guided Melanocyte Detection on Histopathological Images.

Detection of melanocytes serves as a critical prerequisite in assessing melanocytic growth patterns when diagnosing melanoma and its precursor lesions on skin biopsy specimens. However, this detection is challenging due to the visual similarity of melanocytes to other cells in routine Hematoxylin and Eosin (H&E) stained images, leading to the failure of current nuclei detection methods. Stains such as Sox10 can mark melanocytes, but they require an additional step and expense and thus are not regularly used in clinical practice. To address these limitations, we introduce VSGD-Net, a novel detection network that learns melanocyte identification through virtual staining from H&E to Sox10. The method takes only routine H&E images during inference, resulting in a promising approach to support pathologists in the diagnosis of melanoma. To the best of our knowledge, this is the first study that investigates the detection problem using image synthesis features between two distinct pathology stainings. Extensive experimental results show that our proposed model outperforms state-of-the-art nuclei detection methods for melanocyte detection. The source code and pre-trained model are available at: https://github.com/kechunl/VSGD-Net.

Artificial intelligence-enabled mobile electrocardiograms for event prediction in paroxysmal atrial fibrillation.

Background: Paroxysmal atrial fibrillation (AF) often eludes early diagnosis, resulting in significant morbidity and mortality. Artificial intelligence (AI) has been used to predict AF from sinus rhythm electrocardiograms (ECGs), but AF prediction using sinus rhythm mobile electrocardiograms (mECG) remains unexplored. Objective: The purpose of this study was to investigate the utility of AI to predict AF events prospectively and retrospectively using sinus rhythm mECG data. Methods: We trained a neural network to predict AF events from sinus rhythm mECGs obtained from users of the Alivecor KardiaMobile 6L device. We tested our model on sinus rhythm mECGs within ±0-2 days, ±3-7 days, and ±8-30 days from AF events to determine the optimal screening window. Finally, we tested our model on mECGs from before an AF event to determine whether AF can be predicted prospectively. Results: We included 73,861 users with 267,614 mECGs (mean age 58.14 years; 35% women). Users with paroxysmal AF contributed 60.15% of mECGs. Model performance on the test set comprising control and study samples across all windows of interest showed an area under the curve (AUC) score of 0.760 (95% confidence interval [CI] 0.759-0.760), sensitivity of 0.703 (95% CI 0.700-0.705), specificity of 0.684 (95% CI 0.678-0.685), and accuracy of 69.4% (95% CI 0.692-0.700). Model performance was better on ±0-2 day samples (sensitivity 0.711; 95% CI 0.709-0.713) and worse on the ±8-30 day window (sensitivity 0.688; 95% CI 0.685-0.690), with performance on the ±3-7 day window falling in between (sensitivity 0.708; 95% CI 0.704-0.710). Conclusion: Neural networks can predict AF using a widely scalable and cost-effective mobile technology prospectively and retrospectively.

Quilt-1M: One Million Image-Text Pairs for Histopathology.

Recent accelerations in multi-modal applications have been made possible with the plethora of image and text data available online. However, the scarcity of analogous data in the medical field, specifically in histopathology, has halted comparable progress. To enable similar representation learning for histopathology, we turn to YouTube, an untapped resource of videos, offering 1,087 hours of valuable educational histopathology videos from expert clinicians. From YouTube, we curate Quilt: a large-scale vision-language dataset consisting of 768,826 image and text pairs. Quilt was automatically curated using a mixture of models, including large language models, handcrafted algorithms, human knowledge databases, and automatic speech recognition. In comparison, the most comprehensive datasets curated for histopathology amass only around 200K samples. We combine Quilt with datasets from other sources, including Twitter, research papers, and the internet in general, to create an even larger dataset: Quilt-1M, with 1M paired image-text samples, marking it as the largest vision-language histopathology dataset to date. We demonstrate the value of Quilt-1M by fine-tuning a pre-trained CLIP model. Our model outperforms state-of-the-art models on both zero-shot and linear probing tasks for classifying new histopathology images across 13 diverse patch-level datasets of 8 different sub-pathologies and cross-modal retrieval tasks.

Automated analysis of whole slide digital skin biopsy images.

A rapidly increasing rate of melanoma diagnosis has been noted over the past three decades, and nearly 1 in 4 skin biopsies are diagnosed as melanocytic lesions. The gold standard for diagnosis of melanoma is the histopathological examination by a pathologist to analyze biopsy material at both the cellular and structural levels. A pathologist's diagnosis is often subjective and prone to variability, while deep learning image analysis methods may improve and complement current diagnostic and prognostic capabilities. Mitoses are important entities when reviewing skin biopsy cases as their presence carries prognostic information; thus, their precise detection is an important factor for clinical care. In addition, semantic segmentation of clinically important structures in skin biopsies might help the diagnosis pipeline with an accurate classification. We aim to provide prognostic and diagnostic information on skin biopsy images, including the detection of cellular level entities, segmentation of clinically important tissue structures, and other important factors toward the accurate diagnosis of skin biopsy images. This paper is an overview of our work on analysis of digital whole slide skin biopsy images, including mitotic figure (mitosis) detection, semantic segmentation, diagnosis, and analysis of pathologists' viewing patterns, and with new work on melanocyte detection. Deep learning has been applied to our methods for all the detection, segmentation, and diagnosis work. In our studies, deep learning is proven superior to prior approaches to skin biopsy analysis. Our work on analysis of pathologists' viewing patterns is the only such work in the skin biopsy literature. Our work covers the whole spectrum from low-level entities through diagnosis and understanding what pathologists do in performing their diagnoses.

An analysis of pathologists' viewing processes as they diagnose whole slide digital images.

Although pathologists have their own viewing habits while diagnosing, viewing behaviors leading to the most accurate diagnoses are under-investigated. Digital whole slide imaging has enabled investigators to analyze pathologists' visual interpretation of histopathological features using mouse and viewport tracking techniques. In this study, we provide definitions for basic viewing behavior variables and investigate the association of pathologists' characteristics and viewing behaviors, and how they relate to diagnostic accuracy when interpreting whole slide images. We use recordings of 32 pathologists' actions while interpreting a set of 36 digital whole slide skin biopsy images (5 sets of 36 cases; 180 cases total). These viewport tracking data include the coordinates of a viewport scene on pathologists' screens, the magnification level at which that viewport was viewed, as well as a timestamp. We define a set of variables to quantify pathologists' viewing behaviors such as zooming, panning, and interacting with a consensus reference panel's selected region of interest (ROI). We examine the association of these viewing behaviors with pathologists' demographics, clinical characteristics, and diagnostic accuracy using cross-classified multilevel models. Viewing behaviors differ based on clinical experience of the pathologists. Pathologists with a higher caseload of melanocytic skin biopsy cases and pathologists with board certification and/or fellowship training in dermatopathology have lower average zoom and lower variance of zoom levels. Viewing behaviors associated with higher diagnostic accuracy include higher average and variance of zoom levels, a lower magnification percentage (a measure of consecutive zooming behavior), higher total interpretation time, and higher amount of time spent viewing ROIs. Scanning behavior, which refers to panning with a fixed zoom level, has marginally significant positive association with accuracy. Pathologists' training, clinical experience, and their exposure to a range of cases are associated with their viewing behaviors, which may contribute to their diagnostic accuracy. Research in computational pathology integrating digital imaging and clinical informatics opens up new avenues for leveraging viewing behaviors in medical education and training, potentially improving patient care and the effectiveness of clinical workflow.

Improving the Diagnosis of Skin Biopsies Using Tissue Segmentation.

Invasive melanoma, a common type of skin cancer, is considered one of the deadliest. Pathologists routinely evaluate melanocytic lesions to determine the amount of atypia, and if the lesion represents an invasive melanoma, its stage. However, due to the complicated nature of these assessments, inter- and intra-observer variability among pathologists in their interpretation are very common. Machine-learning techniques have shown impressive and robust performance on various tasks including healthcare. In this work, we study the potential of including semantic segmentation of clinically important tissue structure in improving the diagnosis of skin biopsy images. Our experimental results show a 6% improvement in F-score when using whole slide images along with epidermal nests and cancerous dermal nest segmentation masks compared to using whole-slide images alone in training and testing the diagnosis pipeline.

Automated quantitative assessment of amorphous calcifications: Towards improved malignancy risk stratification.

BACKGROUND: Amorphous calcifications noted on mammograms (i.e., small and indistinct calcifications that are difficult to characterize) are associated with high diagnostic uncertainty, often leading to biopsies. Yet, only 20% of biopsied amorphous calcifications are cancer. We present a quantitative approach for distinguishing between benign and actionable (high-risk and malignant) amorphous calcifications using a combination of local textures, global spatial relationships, and interpretable handcrafted expert features. METHOD: Our approach was trained and validated on a set of 168 2D full-field digital mammography exams (248 images) from 168 patients. Within these 248 images, we identified 276 image regions with segmented amorphous calcifications and a biopsy-confirmed diagnosis. A set of local (radiomic and region measurements) and global features (distribution and expert-defined) were extracted from each image. Local features were grouped using an unsupervised k-means clustering algorithm. All global features were concatenated with clustered local features and used to train a LightGBM classifier to distinguish benign from actionable cases. RESULTS: On the held-out test set of 60 images, our approach achieved a sensitivity of 100%, specificity of 35%, and a positive predictive value of 38% when the decision threshold was set to 0.4. Given that all of the images in our test set resulted in a recommendation of a biopsy, the use of our algorithm would have identified 15 images (25%) that were benign, potentially reducing the number of breast biopsies. CONCLUSIONS: Quantitative analysis of full-field digital mammograms can extract subtle shape, texture, and distribution features that may help to distinguish between benign and actionable amorphous calcifications.

End-to-End diagnosis of breast biopsy images with transformers.

Diagnostic disagreements among pathologists occur throughout the spectrum of benign to malignant lesions. A computer-aided diagnostic system capable of reducing uncertainties would have important clinical impact. To develop a computer-aided diagnosis method for classifying breast biopsy images into a range of diagnostic categories (benign, atypia, ductal carcinoma in situ, and invasive breast cancer), we introduce a transformer-based hollistic attention network called HATNet. Unlike state-of-the-art histopathological image classification systems that use a two pronged approach, i.e., they first learn local representations using a multi-instance learning framework and then combine these local representations to produce image-level decisions, HATNet streamlines the histopathological image classification pipeline and shows how to learn representations from gigapixel size images end-to-end. HATNet extends the bag-of-words approach and uses self-attention to encode global information, allowing it to learn representations from clinically relevant tissue structures without any explicit supervision. It outperforms the previous best network Y-Net, which uses supervision in the form of tissue-level segmentation masks, by 8%. Importantly, our analysis reveals that HATNet learns representations from clinically relevant structures, and it matches the classification accuracy of 87 U.S. pathologists for this challenging test set.

Segmenting Skin Biopsy Images with Coarse and Sparse Annotations using U-Net.

The number of melanoma diagnoses has increased dramatically over the past three decades, outpacing almost all other cancers. Nearly 1 in 4 skin biopsies is of melanocytic lesions, highlighting the clinical and public health importance of correct diagnosis. Deep learning image analysis methods may improve and complement current diagnostic and prognostic capabilities. The histologic evaluation of melanocytic lesions, including melanoma and its precursors, involves determining whether the melanocytic population involves the epidermis, dermis, or both. Semantic segmentation of clinically important structures in skin biopsies is a crucial step towards an accurate diagnosis. While training a segmentation model requires ground-truth labels, annotation of large images is a labor-intensive task. This issue becomes especially pronounced in a medical image dataset in which expert annotation is the gold standard. In this paper, we propose a two-stage segmentation pipeline using coarse and sparse annotations on a small region of the whole slide image as the training set. Segmentation results on whole slide images show promising performance for the proposed pipeline.

Brain-Aware Replacements for Supervised Contrastive Learning in Detection of Alzheimer's Disease.

We propose a novel framework for Alzheimer's disease (AD) detection using brain MRIs. The framework starts with a data augmentation method called Brain-Aware Replacements (BAR), which leverages a standard brain parcellation to replace medically-relevant 3D brain regions in an anchor MRI from a randomly picked MRI to create synthetic samples. Ground truth "hard" labels are also linearly mixed depending on the replacement ratio in order to create "soft" labels. BAR produces a great variety of realistic-looking synthetic MRIs with higher local variability compared to other mix-based methods, such as CutMix. On top of BAR, we propose using a soft-label-capable supervised contrastive loss, aiming to learn the relative similarity of representations that reflect how mixed are the synthetic MRIs using our soft labels. This way, we do not fully exhaust the entropic capacity of our hard labels, since we only use them to create soft labels and synthetic MRIs through BAR. We show that a model pre-trained using our framework can be further fine-tuned with a cross-entropy loss using the hard labels that were used to create the synthetic samples. We validated the performance of our framework in a binary AD detection task against both from-scratch supervised training and state-of-the-art self-supervised training plus fine-tuning approaches. Then we evaluated BAR's individual performance compared to another mix-based method CutMix by integrating it within our framework. We show that our framework yields superior results in both precision and recall for the AD detection task.

Machine learning modeling of genome-wide copy number alteration signatures reliably predicts IDH mutational status in adult diffuse glioma.

Knowledge of 1p/19q-codeletion and IDH1/2 mutational status is necessary to interpret any investigational study of diffuse gliomas in the modern era. While DNA sequencing is the gold standard for determining IDH mutational status, genome-wide methylation arrays and gene expression profiling have been used for surrogate mutational determination. Previous studies by our group suggest that 1p/19q-codeletion and IDH mutational status can be predicted by genome-wide somatic copy number alteration (SCNA) data alone, however a rigorous model to accomplish this task has yet to be established. In this study, we used SCNA data from 786 adult diffuse gliomas in The Cancer Genome Atlas (TCGA) to develop a two-stage classification system that identifies 1p/19q-codeleted oligodendrogliomas and predicts the IDH mutational status of astrocytic tumors using a machine-learning model. Cross-validated results on TCGA SCNA data showed near perfect classification results. Furthermore, our astrocytic IDH mutation model validated well on four additional datasets (AUC = 0.97, AUC = 0.99, AUC = 0.95, AUC = 0.96) as did our 1p/19q-codeleted oligodendroglioma screen on the two datasets that contained oligodendrogliomas (MCC = 0.97, MCC = 0.97). We then retrained our system using data from these validation sets and applied our system to a cohort of REMBRANDT study subjects for whom SCNA data, but not IDH mutational status, is available. Overall, using genome-wide SCNAs, we successfully developed a system to robustly predict 1p/19q-codeletion and IDH mutational status in diffuse gliomas. This system can assign molecular subtype labels to tumor samples of retrospective diffuse glioma cohorts that lack 1p/19q-codeletion and IDH mutational status, such as the REMBRANDT study, recasting these datasets as validation cohorts for diffuse glioma research.

Proteoglycan 4 (PRG4) expression and function in dry eye associated inflammation.

Dry eye disease (DED) affects hundreds of millions of people worldwide. It is characterized by the production of inflammatory cytokines and chemokines as well as damaging matrix metalloproteinases (MMPs) at the ocular surface. While proteoglycan 4 (PRG4), a mucin-like glycoprotein present at the ocular surface, is most well known as a boundary lubricant that contributes to ocular surface integrity, it has been shown to blunt inflammation in various cell types, suggesting a dual mechanism of action. Recently, full-length recombinant human PRG4 (rhPRG4) has been shown to improve signs and symptoms of DED in humans. However, there remains a significant need for basic science research on rhPRG4's biological properties and its potential therapeutic mechanisms of action in treating DED. Therefore, the objectives of this study were to characterize endogenous PRG4 expression by telomerase-immortalized human corneal epithelial (hTCEpi) cells, examine whether exogenous rhPRG4 modulates cytokine and chemokine secretion in response to dry eye associated inflammation (TNFα and IL-1ß), explore interactions between rhPRG4 and MMP-9, and understand how experimental dry eye (EDE) in mice affects PRG4 expression. PRG4 secretion from hTCEpi cells was quantified by Western blot and expression visualized by immunocytochemistry. Cytokine/chemokine production was measured by ELISA and Luminex, while rhPRG4's effect on MMP-9 activity, binding, and expression was quantified using an MMP-9 inhibitor kit, surface plasmon resonance, and reverse transcription polymerase chain reaction (RT-PCR), respectively. Finally, EDE was induced in mice, and PRG4 was visualized by immunohistochemistry in the cornea and by Western blot in lacrimal gland lysate. In vitro results demonstrate that hTCEpi cells synthesize and secrete PRG4, and PRG4 secretion is inhibited by TNFα and IL-1ß. In response to these pro-inflammatory stresses, exogenous rhPRG4 significantly reduced the stimulated production of IP-10, RANTES, ENA-78, GROα, MIP-3α, and MIG, and trended towards a reduction of MIP-1α and MIP-1ß. The hTCEpi cells were also able to internalize fluorescently-labelled rhPRG4, consistent with a mechanism of action that includes downstream biological signaling pathways. rhPRG4 was not digested by MMP-9, and it did not modulate MMP-9 gene expression in hTCEpi cells, but it was able to bind to MMP-9 and inhibited in vitro activity of exogenous MMP-9 in the presence of human tears. Finally, in vivo results demonstrate that EDE significantly decreased immunolocalization of PRG4 on the corneal epithelium and trended towards a reduction of PRG4 in lacrimal gland lysate. Collectively these results demonstrate rhPRG4 has anti-inflammatory properties on corneal epithelial cells, particularly as it relates to mitigating chemokine production, and is an inhibitor of MMP-9 activity, as well as that in vivo expression of PRG4 can be altered in preclinical models of DED. In conclusion, these findings contribute to our understanding of PRG4's immunomodulatory properties in the context of DED inflammation and provide the foundation and motivation for further mechanistic research of PRG4's properties on the ocular surface as well as expanding clinical evaluation of its ability as a multifunctional therapeutic agent to effectively provide relief to those who suffer from DED.

CD13 is a critical regulator of cell-cell fusion in osteoclastogenesis.

The transmembrane aminopeptidase CD13 is highly expressed in cells of the myeloid lineage, regulates dynamin-dependent receptor endocytosis and recycling and is a necessary component of actin cytoskeletal organization. Here, we show that CD13-deficient mice present a low bone density phenotype with increased numbers of osteoclasts per bone surface, but display a normal distribution of osteoclast progenitor populations in the bone marrow and periphery. In addition, the bone formation and mineral apposition rates are similar between genotypes, indicating a defect in osteoclast-specific function in vivo. Lack of CD13 led to exaggerated in vitro osteoclastogenesis as indicated by significantly enhanced fusion of bone marrow-derived multinucleated osteoclasts in the presence of M-CSF and RANKL, resulting in abnormally large cells containing remarkably high numbers of nuclei. Mechanistically, while expression levels of the fusion-regulatory proteins dynamin and DC-STAMP1 must be downregulated for fusion to proceed, these are aberrantly sustained at high levels even in CD13-deficient mature multi-nucleated osteoclasts. Further, the stability of fusion-promoting proteins is maintained in the absence of CD13, implicating CD13 in protein turnover mechanisms. Together, we conclude that CD13 may regulate cell-cell fusion by controlling the expression and localization of key fusion regulatory proteins that are critical for osteoclast fusion.

Radiogenomic modeling predicts survival-associated prognostic groups in glioblastoma.

BACKGROUND: Combined whole-exome sequencing (WES) and somatic copy number alteration (SCNA) information can separate isocitrate dehydrogenase (IDH)1/2-wildtype glioblastoma into two prognostic molecular subtypes, which cannot be distinguished by epigenetic or clinical features. The potential for radiographic features to discriminate between these molecular subtypes has yet to be established. METHODS: Radiologic features (n = 35 340) were extracted from 46 multisequence, pre-operative magnetic resonance imaging (MRI) scans of IDH1/2-wildtype glioblastoma patients from The Cancer Imaging Archive (TCIA), all of whom have corresponding WES/SCNA data. We developed a novel feature selection method that leverages the structure of extracted MRI features to mitigate the dimensionality challenge posed by the disparity between a large number of features and the limited patients in our cohort. Six traditional machine learning classifiers were trained to distinguish molecular subtypes using our feature selection method, which was compared to least absolute shrinkage and selection operator (LASSO) feature selection, recursive feature elimination, and variance thresholding. RESULTS: We were able to classify glioblastomas into two prognostic subgroups with a cross-validated area under the curve score of 0.80 (±0.03) using ridge logistic regression on the 15-dimensional principle component analysis (PCA) embedding of the features selected by our novel feature selection method. An interrogation of the selected features suggested that features describing contours in the T2 signal abnormality region on the T2-weighted fluid-attenuated inversion recovery (FLAIR) MRI sequence may best distinguish these two groups from one another. CONCLUSIONS: We successfully trained a machine learning model that allows for relevant targeted feature extraction from standard MRI to accurately predict molecularly-defined risk-stratifying IDH1/2-wildtype glioblastoma patient groups.

Scale-Aware Transformers for Diagnosing Melanocytic Lesions.

Diagnosing melanocytic lesions is one of the most challenging areas of pathology with extensive intra- and inter-observer variability. The gold standard for a diagnosis of invasive melanoma is the examination of histopathological whole slide skin biopsy images by an experienced dermatopathologist. Digitized whole slide images offer novel opportunities for computer programs to improve the diagnostic performance of pathologists. In order to automatically classify such images, representations that reflect the content and context of the input images are needed. In this paper, we introduce a novel self-attention-based network to learn representations from digital whole slide images of melanocytic skin lesions at multiple scales. Our model softly weighs representations from multiple scales, allowing it to discriminate between diagnosis-relevant and -irrelevant information automatically. Our experiments show that our method outperforms five other state-of-the-art whole slide image classification methods by a significant margin. Our method also achieves comparable performance to 187 practicing U.S. pathologists who interpreted the same cases in an independent study. To facilitate relevant research, full training and inference code is made publicly available at https://github.com/meredith-wenjunwu/ScATNet.

Analysis of Regions of Interest and Distractor Regions in Breast Biopsy Images.

This paper studies why pathologists can misdiagnose diagnostically challenging breast biopsy cases, using a data set of 240 whole slide images (WSIs). Three experienced pathologists agreed on a consensus reference ground-truth diagnosis for each slide and also a consensus region of interest (ROI) from which the diagnosis could best be made. A study group of 87 other pathologists then diagnosed test sets (60 slides each) and marked their own regions of interest. Diagnoses and ROIs were categorized such that if on a given slide, their ROI differed from the consensus ROI and their diagnosis was incorrect, that ROI was called a distractor. We used the HATNet transformer-based deep learning classifier to evaluate the visual similarities and differences between the true (consensus) ROIs and the distractors. Results showed high accuracy for both the similarity and difference networks, showcasing the challenging nature of feature classification with breast biopsy images. This study is important in the potential use of its results for teaching pathologists how to diagnose breast biopsy slides.