Testing of rapid evaporative mass spectrometry for histological tissue classification and molecular diagnostics in a multi-site study.

BACKGROUND: While REIMS technology has successfully been demonstrated for the histological identification of ex-vivo breast tumor tissues, questions regarding the robustness of the approach and the possibility of tumor molecular diagnostics still remain unanswered. In the current study, we set out to determine whether it is possible to acquire cross-comparable REIMS datasets at multiple sites for the identification of breast tumors and subtypes. METHODS: A consortium of four sites with three of them having access to fresh surgical tissue samples performed tissue analysis using identical REIMS setups and protocols. Overall, 21 breast cancer specimens containing pathology-validated tumor and adipose tissues were analyzed and results were compared using uni- and multivariate statistics on normal, WT and PIK3CA mutant ductal carcinomas. RESULTS: Statistical analysis of data from standards showed significant differences between sites and individual users. However, the multivariate classification models created from breast cancer data elicited 97.1% and 98.6% correct classification for leave-one-site-out and leave-one-patient-out cross validation. Molecular subtypes represented by PIK3CA mutation gave consistent results across sites. CONCLUSIONS: The results clearly demonstrate the feasibility of creating and using global classification models for a REIMS-based margin assessment tool, supporting the clinical translatability of the approach.

ImSpect: Image-driven self-supervised learning for surgical margin evaluation with mass spectrometry.

PURPOSE: Real-time assessment of surgical margins is critical for favorable outcomes in cancer patients. The iKnife is a mass spectrometry device that has demonstrated potential for margin detection in cancer surgery. Previous studies have shown that using deep learning on iKnife data can facilitate real-time tissue characterization. However, none of the existing literature on the iKnife facilitate the use of publicly available, state-of-the-art pretrained networks or datasets that have been used in computer vision and other domains. METHODS: In a new framework we call ImSpect, we convert 1D iKnife data, captured during basal cell carcinoma (BCC) surgery, into 2D images in order to capitalize on state-of-the-art image classification networks. We also use self-supervision to leverage large amounts of unlabeled, intraoperative data to accommodate the data requirements of these networks. RESULTS: Through extensive ablation studies, we show that we can surpass previous benchmarks of margin evaluation in BCC surgery using iKnife data, achieving an area under the receiver operating characteristic curve (AUC) of 81%. We also depict the attention maps of the developed DL models to evaluate the biological relevance of the embedding space CONCLUSIONS: We propose a new method for characterizing tissue at the surgical margins, using mass spectrometry data from cancer surgery.

From quantitative metrics to clinical success: assessing the utility of deep learning for tumor segmentation in breast surgery.

PURPOSE: Preventing positive margins is essential for ensuring favorable patient outcomes following breast-conserving surgery (BCS). Deep learning has the potential to enable this by automatically contouring the tumor and guiding resection in real time. However, evaluation of such models with respect to pathology outcomes is necessary for their successful translation into clinical practice. METHODS: Sixteen deep learning models based on established architectures in the literature are trained on 7318 ultrasound images from 33 patients. Models are ranked by an expert based on their contours generated from images in our test set. Generated contours from each model are also analyzed using recorded cautery trajectories of five navigated BCS cases to predict margin status. Predicted margins are compared with pathology reports. RESULTS: The best-performing model using both quantitative evaluation and our visual ranking framework achieved a mean Dice score of 0.959. Quantitative metrics are positively associated with expert visual rankings. However, the predictive value of generated contours was limited with a sensitivity of 0.750 and a specificity of 0.433 when tested against pathology reports. CONCLUSION: We present a clinical evaluation of deep learning models trained for intraoperative tumor segmentation in breast-conserving surgery. We demonstrate that automatic contouring is limited in predicting pathology margins despite achieving high performance on quantitative metrics.

Preoperative CT and survival data for patients undergoing resection of colorectal liver metastases.

The liver is a common site for the development of metastases in colorectal cancer. Treatment selection for patients with colorectal liver metastases (CRLM) is difficult; although hepatic resection will cure a minority of CRLM patients, recurrence is common. Reliable preoperative prediction of recurrence could therefore be a valuable tool for physicians in selecting the best candidates for hepatic resection in the treatment of CRLM. It has been hypothesized that evidence for recurrence could be found via quantitative image analysis on preoperative CT imaging of the future liver remnant before resection. To investigate this hypothesis, we have collected preoperative hepatic CT scans, clinicopathologic data, and recurrence/survival data, from a large, single-institution series of patients (n = 197) who underwent hepatic resection of CRLM. For each patient, we also created segmentations of the liver, vessels, tumors, and future liver remnant. The largest of its kind, this dataset is a resource that may aid in the development of quantitative imaging biomarkers and machine learning models for the prediction of post-resection hepatic recurrence of CRLM.

Real-time integration between Microsoft HoloLens 2 and 3D Slicer with demonstration in pedicle screw placement planning.

PURPOSE: Up to date, there has been a lack of software infrastructure to connect 3D Slicer to any augmented reality (AR) device. This work describes a novel connection approach using Microsoft HoloLens 2 and OpenIGTLink, with a demonstration in pedicle screw placement planning. METHODS: We developed an AR application in Unity that is wirelessly rendered onto Microsoft HoloLens 2 using Holographic Remoting. Simultaneously, Unity connects to 3D Slicer using the OpenIGTLink communication protocol. Geometrical transform and image messages are transferred between both platforms in real time. Through the AR glasses, a user visualizes a patient's computed tomography overlaid onto virtual 3D models showing anatomical structures. We technically evaluated the system by measuring message transference latency between the platforms. Its functionality was assessed in pedicle screw placement planning. Six volunteers planned pedicle screws' position and orientation with the AR system and on a 2D desktop planner. We compared the placement accuracy of each screw with both methods. Finally, we administered a questionnaire to all participants to assess their experience with the AR system. RESULTS: The latency in message exchange is sufficiently low to enable real-time communication between the platforms. The AR method was non-inferior to the 2D desktop planner, with a mean error of 2.1 ± 1.4 mm. Moreover, 98% of the screw placements performed with the AR system were successful, according to the Gertzbein-Robbins scale. The average questionnaire outcomes were 4.5/5. CONCLUSIONS: Real-time communication between Microsoft HoloLens 2 and 3D Slicer is feasible and supports accurate planning for pedicle screw placement.

Development and validation of a radiofrequency ablation treatment planning system for vertebral metastases.

PURPOSE: Bone-targeted radiofrequency ablation (RFA) is widely used in the treatment of vertebral metastases. While radiation therapy utilizes established treatment planning systems (TPS) based on multimodal imaging to optimize treatment volumes, current RFA of vertebral metastases has been limited to qualitative image-based assessment of tumour location to direct probe selection and access. This study aimed to design, develop and evaluate a computational patient-specific RFA TPS for vertebral metastases. METHODS: A TPS was developed on the open-source 3D slicer platform, including procedural setup, dose calculation (based on finite element modelling), and analysis/visualization modules. Usability testing was carried out by 7 clinicians involved in the treatment of vertebral metastases on retrospective clinical imaging data using a simplified dose calculation engine. In vivo evaluation was performed in a preclinical porcine model (n = 6 vertebrae). RESULTS: Dose analysis was successfully performed, with generation and display of thermal dose volumes, thermal damage, dose volume histograms and isodose contours. Usability testing showed an overall positive response to the TPS as beneficial to safe and effective RFA. The in vivo porcine study showed good agreement between the manually segmented thermally damaged volumes vs. the damage volumes identified from the TPS (Dice Similarity Coefficient = 0.71 ± 0.03, Hausdorff distance = 1.2 ± 0.1 mm). CONCLUSION: A TPS specifically dedicated to RFA in the bony spine could help account for tissue heterogeneities in both thermal and electrical properties. A TPS would enable visualization of damage volumes in 2D and 3D, assisting clinicians in decisions about potential safety and effectiveness prior to performing RFA in the metastatic spine.

Metabolically Active Zones Involving Fatty Acid Elongation Delineated by DESI-MSI Correlate with Pathological and Prognostic Features of Colorectal Cancer.

Colorectal cancer (CRC) is the second leading cause of cancer deaths. Despite recent advances, five-year survival rates remain largely unchanged. Desorption electrospray ionization mass spectrometry imaging (DESI) is an emerging nondestructive metabolomics-based method that retains the spatial orientation of small-molecule profiles on tissue sections, which may be validated by 'gold standard' histopathology. In this study, CRC samples were analyzed by DESI from 10 patients undergoing surgery at Kingston Health Sciences Center. The spatial correlation of the mass spectral profiles was compared with histopathological annotations and prognostic biomarkers. Fresh frozen sections of representative colorectal cross sections and simulated endoscopic biopsy samples containing tumour and non-neoplastic mucosa for each patient were generated and analyzed by DESI in a blinded fashion. Sections were then hematoxylin and eosin (H and E) stained, annotated by two independent pathologists, and analyzed. Using PCA/LDA-based models, DESI profiles of the cross sections and biopsies achieved 97% and 75% accuracies in identifying the presence of adenocarcinoma, using leave-one-patient-out cross validation. Among the m/z ratios exhibiting the greatest differential abundance in adenocarcinoma were a series of eight long-chain or very-long-chain fatty acids, consistent with molecular and targeted metabolomics indicators of de novo lipogenesis in CRC tissue. Sample stratification based on the presence of lympovascular invasion (LVI), a poor CRC prognostic indicator, revealed the abundance of oxidized phospholipids, suggestive of pro-apoptotic mechanisms, was increased in LVI-negative compared to LVI-positive patients. This study provides evidence of the potential clinical utility of spatially-resolved DESI profiles to enhance the information available to clinicians for CRC diagnosis and prognosis.

nnUNet-based Multi-modality Breast MRI Segmentation and Tissue-Delineating Phantom for Robotic Tumor Surgery Planning.

Segmentation of the thoracic region and breast tissues is crucial for analyzing and diagnosing the presence of breast masses. This paper introduces a medical image segmentation architecture that aggregates two neural networks based on the state-of-the-art nnU-Net. Additionally, this study proposes a polyvinyl alcohol cryogel (PVA-C) breast phantom, based on its automated segmentation approach, to enable planning and navigation experiments for robotic breast surgery. The dataset consists of multimodality breast MRI of T2W and STIR images obtained from 10 patients. A statistical analysis of segmentation tasks emphasizes the Dice Similarity Coefficient (DSC), segmentation accuracy, sensitivity, and specificity. We first use a single class labeling to segment the breast region and then exploit it as an input for three-class labeling to segment fatty, fibroglandular (FGT), and tumorous tissues. The first network has a 0.95 DCS, while the second network has a 0.95, 0.83, and 0.41 for fat, FGT, and tumor classes, respectively. Clinical Relevance-This research is relevant to the breast surgery community as it establishes a deep learning-based (DL) algorithmic and phantomic foundation for surgical planning and navigation that will exploit preoperative multimodal MRI and intraoperative ultrasound to achieve highly cosmetic breast surgery. In addition, the planning and navigation will guide a robot that can cut, resect, bag, and grasp a tissue mass that encapsulates breast tumors and positive tissue margins. This image-guided robotic approach promises to potentiate the accuracy of breast surgeons and improve patient outcomes.

Uncertainty estimation for margin detection in cancer surgery using mass spectrometry.

PURPOSE: Rapid evaporative ionization mass spectrometry (REIMS) is an emerging technology for clinical margin detection. Deployment of REIMS depends on construction of reliable deep learning models that can categorize tissue according to its metabolomic signature. Challenges associated with developing these models include the presence of noise during data acquisition and the variance in tissue signatures between patients. In this study, we propose integration of uncertainty estimation in deep models to factor predictive confidence into margin detection in cancer surgery. METHODS: iKnife is used to collect 693 spectra of cancer and healthy samples acquired from 91 patients during basal cell carcinoma resection. A Bayesian neural network and two baseline models are trained on these data to perform classification as well as uncertainty estimation. The samples with high estimated uncertainty are then removed, and new models are trained using the clean data. The performance of proposed and baseline models, with different ratios of filtered data, is then compared. RESULTS: The data filtering does not improve the performance of the baseline models as they cannot provide reliable estimations of uncertainty. In comparison, the proposed model demonstrates a statistically significant improvement in average balanced accuracy (75.2%), sensitivity (74.1%) and AUC (82.1%) after removing uncertain training samples. We also demonstrate that if highly uncertain samples are predicted and removed from the test data, sensitivity further improves to 88.2%. CONCLUSIONS: This is the first study that applies uncertainty estimation to inform model training and deployment for tissue recognition in cancer surgery. Uncertainty estimation is leveraged in two ways: by factoring a measure of input noise in training the models and by including predictive confidence in reporting the outputs. We empirically show that considering uncertainty for model development can help improve the overall accuracy of a margin detection system using REIMS.

SlicerHeart: An open-source computing platform for cardiac image analysis and modeling.

Cardiovascular disease is a significant cause of morbidity and mortality in the developed world. 3D imaging of the heart's structure is critical to the understanding and treatment of cardiovascular disease. However, open-source tools for image analysis of cardiac images, particularly 3D echocardiographic (3DE) data, are limited. We describe the rationale, development, implementation, and application of SlicerHeart, a cardiac-focused toolkit for image analysis built upon 3D Slicer, an open-source image computing platform. We designed and implemented multiple Python scripted modules within 3D Slicer to import, register, and view 3DE data, including new code to volume render and crop 3DE. In addition, we developed dedicated workflows for the modeling and quantitative analysis of multi-modality image-derived heart models, including heart valves. Finally, we created and integrated new functionality to facilitate the planning of cardiac interventions and surgery. We demonstrate application of SlicerHeart to a diverse range of cardiovascular modeling and simulation including volume rendering of 3DE images, mitral valve modeling, transcatheter device modeling, and planning of complex surgical intervention such as cardiac baffle creation. SlicerHeart is an evolving open-source image processing platform based on 3D Slicer initiated to support the investigation and treatment of congenital heart disease. The technology in SlicerHeart provides a robust foundation for 3D image-based investigation in cardiovascular medicine.

Sensor-Based Automated Detection of Electrosurgical Cautery States.

In computer-assisted surgery, it is typically required to detect when the tool comes into contact with the patient. In activated electrosurgery, this is known as the energy event. By continuously tracking the electrosurgical tools' location using a navigation system, energy events can help determine locations of sensor-classified tissues. Our objective was to detect the energy event and determine the settings of electrosurgical cautery-robustly and automatically based on sensor data. This study aims to demonstrate the feasibility of using the cautery state to detect surgical incisions, without disrupting the surgical workflow. We detected current changes in the wires of the cautery device and grounding pad using non-invasive current sensors and an oscilloscope. An open-source software was implemented to apply machine learning on sensor data to detect energy events and cautery settings. Our methods classified each cautery state at an average accuracy of 95.56% across different tissue types and energy level parameters altered by surgeons during an operation. Our results demonstrate the feasibility of automatically identifying energy events during surgical incisions, which could be an important safety feature in robotic and computer-integrated surgery. This study provides a key step towards locating tissue classifications during breast cancer operations and reducing the rate of positive margins.

Real-time automatic tumor segmentation for ultrasound-guided breast-conserving surgery navigation.

PURPOSE: Ultrasound-based navigation is a promising method in breast-conserving surgery, but tumor contouring often requires a radiologist at the time of surgery. Our goal is to develop a real-time automatic neural network-based tumor contouring process for intraoperative guidance. Segmentation accuracy is evaluated by both pixel-based metrics and expert visual rating. METHODS: This retrospective study includes 7318 intraoperative ultrasound images acquired from 33 breast cancer patients, randomly split between 80:20 for training and testing. We implement a u-net architecture to label each pixel on ultrasound images as either tumor or healthy breast tissue. Quantitative metrics are calculated to evaluate the model's accuracy. Contour quality and usability are also assessed by fellowship-trained breast radiologists and surgical oncologists. Additionally, the viability of using our u-net model in an existing surgical navigation system is evaluated by measuring the segmentation frame rate. RESULTS: The mean dice similarity coefficient of our u-net model is 0.78, with an area under the receiver-operating characteristics curve of 0.94, sensitivity of 0.95, and specificity of 0.67. Expert visual ratings are positive, with 93% of responses rating tumor contour quality at or above 7/10, and 75% of responses rating contour quality at or above 8/10. Real-time tumor segmentation achieved a frame rate of 16 frames-per-second, sufficient for clinical use. CONCLUSION: Neural networks trained with intraoperative ultrasound images provide consistent tumor segmentations that are well received by clinicians. These findings suggest that neural networks are a promising adjunct to alleviate radiologist workload as well as improving efficiency in breast-conserving surgery navigation systems.

Application of Intraoperative Mass Spectrometry and Data Analytics for Oncological Margin Detection, A Review.

OBJECTIVE: A common phase of early-stage oncological treatment is the surgical resection of cancerous tissue. The presence of cancer cells on the resection margin, referred to as positive margin, is correlated with the recurrence of cancer and may require re-operation, negatively impacting many facets of patient outcomes. There exists a significant gap in the surgeon's ability to intraoperatively delineate between tissues. Mass spectrometry methods have shown considerable promise as intraoperative tissue profiling tools that can assist with the complete resection of cancer. To do so, the vastness of the information collected through these modalities must be digested, relying on robust and efficient extraction of insights through data analysis pipelines. METHODS: We review clinical mass spectrometry literature and prioritize intraoperatively applied modalities. We also survey the data analysis methods employed in these studies. RESULTS: Our review outlines the advantages and shortcomings of mass spectrometry imaging and point-based tissue probing methods. For each modality, we identify statistical, linear transformation and machine learning techniques that demonstrate high performance in classifying cancerous tissues across several organ systems. A limited number of studies presented results captured intraoperatively. CONCLUSION: Through continued research of data centric techniques, like mass spectrometry, and the development of robust analysis approaches, intraoperative margin assessment is becoming feasible. SIGNIFICANCE: By establishing the relatively short history of mass spectrometry techniques applied to surgical studies, we hope to inform future applications and aid in the selection of suitable data analysis frameworks for the development of intraoperative margin detection technologies.

Dynamic Annular Modeling of the Unrepaired Complete Atrioventricular Canal Annulus.

BACKGROUND: Repair of complete atrioventricular canal (CAVC) is often complicated by atrioventricular valve regurgitation, particularly of the left-sided valve. Understanding the 3-dimensional (3D) structure of the atrioventricular canal annulus before repair may help to inform optimized repair. However, the 3D shape and movement of the CAVC annulus has been neither quantified nor rigorously compared with a normal mitral valve annulus. METHODS: The complete annuli of 43 patients with CAVC were modeled in 4 cardiac phases using transthoracic 3D echocardiograms and custom code. The annular structure was compared with the annuli of 20 normal pediatric mitral valves using 3D metrics and statistical shape analysis (Procrustes analysis). RESULTS: The unrepaired CAVC annulus varied in shape significantly throughout the cardiac cycle. Procrustes analysis visually demonstrated that the average normalized CAVC annular shape is more planar than the normal mitral annulus. Quantitatively, the annular height-to-valve width ratio of the native left CAVC atrioventricular valve was significantly lower than that of a normal mitral valve in all systolic phases (P < .001). CONCLUSIONS: The left half of the CAVC annulus is more planar than that of a normal mitral valve with an annular height-to-valve width ratio similar to dysfunctional mitral valves. Given the known importance of annular shape to mitral valve function, further exploration of the association of 3D structure to valve function in CAVC is warranted.

Training in soft tissue resection using real-time visual computer navigation feedback from the Surgery Tutor: A randomized controlled trial.

BACKGROUND: In competency-based medical education, surgery trainees are often required to learn procedural skills in a simulated setting before proceeding to the clinical environment. The Surgery Tutor computer navigation platform allows for real-time proctor-less assessment of open soft tissue resection skills; however, the use of this platform as an aid in acquisition of procedural skills is yet to be explored. METHODS: In this prospective randomized controlled trial, 20 final year medical students were randomized to receive either training with real-time computer navigation feedback (Intervention, n = 10) or simulation training without navigation feedback (Control, n = 10) during resection of simulated non-palpable soft tissue tumors. Real-time computer navigation feedback allowed participants to visualize the position of their scalpel relative to the tumor. Computer navigation feedback was removed for postintervention assessment. Primary outcome was positive margin rate. Secondary outcomes were procedure time, mass of tissue excised, number of scalpel motions, and distance traveled by the scalpel. RESULTS: Training with real-time computer navigation resulted in a significantly lower positive margin rate as compared to training without navigation feedback (0% vs 40%, P = .025). All other performance metrics were not significantly different between the 2 groups. Participants in the intervention group displayed significant improvement in positive margin rate from baseline to final assessment (80% vs 0%, P < .01), whereas participants in the Control group did not. CONCLUSION: Real-time visual computer navigation feedback from the Surgery Tutor resulted in superior acquisition of procedural skills as compared to training without navigation feedback.

Surgical data science - from concepts toward clinical translation.

Recent developments in data science in general and machine learning in particular have transformed the way experts envision the future of surgery. Surgical Data Science (SDS) is a new research field that aims to improve the quality of interventional healthcare through the capture, organization, analysis and modeling of data. While an increasing number of data-driven approaches and clinical applications have been studied in the fields of radiological and clinical data science, translational success stories are still lacking in surgery. In this publication, we shed light on the underlying reasons and provide a roadmap for future advances in the field. Based on an international workshop involving leading researchers in the field of SDS, we review current practice, key achievements and initiatives as well as available standards and tools for a number of topics relevant to the field, namely (1) infrastructure for data acquisition, storage and access in the presence of regulatory constraints, (2) data annotation and sharing and (3) data analytics. We further complement this technical perspective with (4) a review of currently available SDS products and the translational progress from academia and (5) a roadmap for faster clinical translation and exploitation of the full potential of SDS, based on an international multi-round Delphi process.

Combined Mass Spectrometry and Histopathology Imaging for Perioperative Tissue Assessment in Cancer Surgery.

Mass spectrometry is an effective imaging tool for evaluating biological tissue to detect cancer. With the assistance of deep learning, this technology can be used as a perioperative tissue assessment tool that will facilitate informed surgical decisions. To achieve such a system requires the development of a database of mass spectrometry signals and their corresponding pathology labels. Assigning correct labels, in turn, necessitates precise spatial registration of histopathology and mass spectrometry data. This is a challenging task due to the domain differences and noisy nature of images. In this study, we create a registration framework for mass spectrometry and pathology images as a contribution to the development of perioperative tissue assessment. In doing so, we explore two opportunities in deep learning for medical image registration, namely, unsupervised, multi-modal deformable image registration and evaluation of the registration. We test this system on prostate needle biopsy cores that were imaged with desorption electrospray ionization mass spectrometry (DESI) and show that we can successfully register DESI and histology images to achieve accurate alignment and, consequently, labelling for future training. This automation is expected to improve the efficiency and development of a deep learning architecture that will benefit the use of mass spectrometry imaging for cancer diagnosis.

Design of an Ultrasound-Navigated Prostate Cancer Biopsy System for Nationwide Implementation in Senegal.

This paper presents the design of NaviPBx, an ultrasound-navigated prostate cancer biopsy system. NaviPBx is designed to support an affordable and sustainable national healthcare program in Senegal. It uses spatiotemporal navigation and multiparametric transrectal ultrasound to guide biopsies. NaviPBx integrates concepts and methods that have been independently validated previously in clinical feasibility studies and deploys them together in a practical prostate cancer biopsy system. NaviPBx is based entirely on free open-source software and will be shared as a free open-source program with no restriction on its use. NaviPBx is set to be deployed and sustained nationwide through the Senegalese Military Health Service. This paper reports on the results of the design process of NaviPBx. Our approach concentrates on "frugal technology", intended to be affordable for low-middle income (LMIC) countries. Our project promises the wide-scale application of prostate biopsy and will foster time-efficient development and programmatic implementation of ultrasound-guided diagnostic and therapeutic interventions in Senegal and beyond.

Lack of definitive presurgical pathological diagnosis is associated with inadequate surgical margins in breast-conserving surgery.

PURPOSE: To determine the impact of definitive presurgical diagnosis on surgical margins in breast-conserving surgery (BCS) for primary carcinomas; clinicopathological features were also analyzed. METHODS: This retrospective study included women who underwent BCS for primary carcinomas in 2016 and 2017. Definitive presurgical diagnosis was defined as having a presurgical core needle biopsy (CNB) and not being upstaged between biopsy and surgery. Biopsy data and imaging findings including breast density were retrieved. Inadequate surgical margins (IM) were defined per latest ASCO and ASTRO guidelines. Univariable and multivariable analyses were performed. RESULTS: 360 women (median age, 66) met inclusion criteria with 1 having 2 cancers. 82.5% (298/361) were invasive cancers while 17.5% (63/361) were ductal carcinoma in situ (DCIS). Most biopsies were US-guided (284/346, 82.0%), followed by mammographic (60/346, 17.3%), and MRI-guided (2/346, 0.6%). US and mammographic CNB yielded median samples of 2 and 4, respectively, with a 14G needle. 15 patients (4.2%) lacked presurgical CNB. The IM rate was 30.0%. In multivariable analysis, large invasive cancers (>20 mm), dense breasts, and DCIS were associated with IM (p = 0.029, p = 0.010, and p = 0.013, respectively). Most importantly, lack of definitive presurgical diagnosis was a risk factor for IM (OR, 2.35; 95% CI: 1.23-4.51, p = 0.010). In contrast, neither patient age (<50) nor aggressive features (e.g., LVI) were associated with IM. CONCLUSION: Lack of a definitive presurgical diagnosis was associated with a two-fold increase of IM in BCS; other risk factors were dense breasts, large invasive cancers, and DCIS.

Domain adaptation and self-supervised learning for surgical margin detection.

PURPOSE: One in five women who undergo breast conserving surgery will need a second revision surgery due to remaining tumor. The iKnife is a mass spectrometry modality that produces real-time margin information based on the metabolite signatures in surgical smoke. Using this modality and real-time tissue classification, surgeons could remove all cancerous tissue during the initial surgery, improving many facets of patient outcomes. An obstacle in developing a iKnife breast cancer recognition model is the destructive, time-consuming and sensitive nature of the data collection that limits the size of the datasets. METHODS: We address these challenges by first, building a self-supervised learning model from limited, weakly labeled data. By doing so, the model can learn to contextualize the general features of iKnife data from a more accessible cancer type. Second, the trained model can then be applied to a cancer classification task on breast data. This domain adaptation allows for the transfer of learnt weights from models of one tissue type to another. RESULTS: Our datasets contained 320 skin burns (129 tumor burns, 191 normal burns) from 51 patients and 144 breast tissue burns (41 tumor and 103 normal) from 11 patients. We investigate the effect of different hyper-parameters on the performance of the final classifier. The proposed two-step method performed statistically significantly better than a baseline model (p-value < 0.0001), by achieving an accuracy, sensitivity and specificity of 92%, 88% and 92%, respectively. CONCLUSION: This is the first application of domain transfer for iKnife REIMS data. We showed that having a limited number of breast data samples for training a classifier can be compensated by self-supervised learning and domain adaption on a set of unlabeled skin data. We plan to confirm this performance by collecting new breast samples and extending it to incorporate other cancer tissues.