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.

Interval Cancer Detection Using a Neural Network and Breast Density in Women with Negative Screening Mammograms.

Background Inclusion of mammographic breast density (BD) in breast cancer risk models improves accuracy, but accuracy remains modest. Interval cancer (IC) risk prediction may be improved by combining assessments of BD and an artificial intelligence (AI) cancer detection system. Purpose To evaluate the performance of a neural network (NN)-based model that combines the assessments of BD and an AI system in the prediction of risk of developing IC among women with negative screening mammography results. Materials and Methods This retrospective nested case-control study performed with screening examinations included women who developed IC and women with normal follow-up findings (from January 2011 to January 2015). An AI cancer detection system analyzed all studies yielding a score of 1-10, representing increasing likelihood of malignancy. BD was automatically computed using publicly available software. An NN model was trained by combining the AI score and BD using 10-fold cross-validation. Bootstrap analysis was used to calculate the area under the receiver operating characteristic curve (AUC), sensitivity at 90% specificity, and 95% CIs of the AI, BD, and NN models. Results A total of 2222 women with IC and 4661 women in the control group were included (mean age, 61 years; age range, 49-76 years). AUC of the NN model was 0.79 (95% CI: 0.77,0.81), which was higher than AUC of the AI cancer detection system or BD alone (AUC, 0.73 [95% CI: 0.71, 0.76] and 0.69 [95% CI: 0.67, 0.71], respectively; P < .001 for both). At 90% specificity, the NN model had a sensitivity of 50.9% (339 of 666 women; 95% CI: 45.2, 56.3) for prediction of IC, which was higher than that of the AI system (37.5%; 250 of 666 women; 95% CI: 33.0, 43.7; P < .001) or BD percentage alone (22.4%; 149 of 666 women; 95% CI: 17.9, 28.5; P < .001). Conclusion The combined assessment of an artificial intelligence detection system and breast density measurements enabled identification of a larger proportion of women who would develop interval cancer compared with either method alone. Published under a CC BY 4.0 license.

Feasibility of Micro-Computed Tomography Imaging for Direct Assessment of Surgical Resection Margins During Breast-Conserving Surgery.

BACKGROUND: To analyze the feasibility and accuracy of micro-computed tomography (micro-CT) for surgical margin assessment in breast excision specimen. MATERIALS AND METHODS: Two data sets of 30 micro-CT scans were retrospectively evaluated for positive resection margins by four observers in two phases, using pathology as a gold standard. Results of phase 1 were evaluated to define micro-CT evaluation guidelines for phase 2. Interobserver agreement was also assessed (kappa). In addition, a prospective study was conducted in which 40 micro-CT scans were directly acquired, reconstructed, and evaluated for positive resection margins by one observer. A suspect positive resection margin on micro-CT was annotated onto the specimen with ink, enabling local validation by pathology. Main outcome measures were accuracy, sensitivity, specificity, and positive predictive value (PPV). RESULTS: Average accuracy, sensitivity, specificity, and PPV for the four observers were 63%, 38%, 70%, and 22%, respectively, in phase 1 and 72%, 40%, 78%, and 26%, respectively, in phase 2. The interobserver agreement was fair [kappa (range), 0.31 (0.12-0.80) in phase 1 and 0.23 (0-0.43) in phase 2]. In the prospective study 70% of the surgical resection margins were correctly evaluated. Ten specimens were annotated for positive resection margins, which correlated with three positive and three close (<1 mm) margins on pathology. Sensitivity, specificity, and PPV were 38%, 78%, and 30%, respectively. CONCLUSIONS: Micro-CT imaging of breast excision specimen has moderate accuracy and considerable interobserver variation for analysis of surgical resection margins. Especially sensitivity and PPV need to be improved before micro-CT-based margin assessment can be introduced in clinical practice.

Neoantigen Targetability in Progressive Advanced Melanoma.

PURPOSE: The availability of (neo)antigens and the infiltration of tumors by (neo)antigen-specific T cells are crucial factors in cancer immunotherapy. In this study, we aimed to investigate the targetability of (neo)antigens in advanced progessive melanoma and explore the potential for continued T-cell-based immunotherapy. EXPERIMENTAL DESIGN: We examined a cohort of eight patients with melanoma who had sequential metastases resected at early and later time points. Antigen-presenting capacity was assessed using IHC and flow cytometry. T-cell infiltration was quantified through multiplex immunofluorescence. Whole-exome and RNA sequencing were conducted to identify neoantigens and assess the expression of neoantigens and tumor-associated antigens. Mass spectrometry was used to evaluate antigen presentation. Tumor recognition by autologous T cells was assessed by coculture assays with cell lines derived from the metastatic lesions. RESULTS: We observed similar T-cell infiltration in paired early and later metastatic (LM) lesions. Although elements of the antigen-presenting machinery were affected in some LM lesions, both the early and later metastasis-derived cell lines were recognized by autologous T cells. At the genomic level, the (neo)antigen landscape was dynamic, but the (neo)antigen load was stable between paired lesions. CONCLUSIONS: Our findings indicate that subsequently isolated tumors from patients with late-stage melanoma retain sufficient antigen-presenting capacity, T-cell infiltration, and a stable (neo)antigen load, allowing recognition of tumor cells by T cells. This indicates a continuous availability of T-cell targets in metastases occurring at different time points and supports further exploration of (neo)antigen-specific T-cell-based therapeutic approaches for advanced melanoma.

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.

Electromagnetic (EM) catheter path tracking in ultrasound-guided brachytherapy of the breast.

PURPOSE: To evaluate a novel navigation system for breast brachytherapy, based on ultrasound (US)-guided catheter needle implantations followed by electromagnetic (EM) tracking of catheter paths. METHODS: Breast phantoms were produced, containing US-visible tumors. Ultrasound was used to localize the tumor pose and volume within the phantom, followed by planning an optimal catheter pattern through the tumor using navigation software. An electromagnetic (EM)-tracked catheter needle was used to insert the catheters in the desired pattern. The inserted catheters were visualized on a post-implant CT, serving as ground truth. Electromagnetic (EM) tracking and reconstruction of the inserted catheter paths were performed by pulling a flexible EM guidewire through each catheter, performed in two clinical brachytherapy suites. The accuracy of EM catheter tracking was evaluated by calculating the Hausdorff distance between the EM-tracked and CT-based catheter paths. The accuracy and clinical feasibility of EM catheter tracking were also evaluated in three breast cancer patients, performed in a separate experiment room. RESULTS: A total of 71 catheter needles were implanted into 12 phantoms using US guidance and EM navigation, in an average ± SD time of 8.1 ± 2.9 min. The accuracy of EM catheter tracking was dependent on the brachytherapy suite: 2.0 ± 1.2 mm in suite 1 and 0.6 ± 0.2 mm in suite 2. EM catheter tracking was successfully performed in three breast brachytherapy patients. Catheter tracking typically took less than 5 min and had an average accuracy of 1.7 ± 0.3 mm. CONCLUSION: Our preliminary results show a potential role for US guidance and EM needle navigation for implantation of catheters for breast brachytherapy. EM catheter tracking can accurately assess the implant geometry in breast brachytherapy patients. This methodology has the potential to evaluate catheter positions directly after the implantation and during the several fractions of the treatment.

Perioperative margin detection in basal cell carcinoma using a deep learning framework: a feasibility study.

PURPOSE: Basal cell carcinoma (BCC) is the most commonly diagnosed cancer and the number of diagnosis is growing worldwide due to increased exposure to solar radiation and the aging population. Reduction of positive margin rates when removing BCC leads to fewer revision surgeries and consequently lower health care costs, improved cosmetic outcomes and better patient care. In this study, we propose the first use of a perioperative mass spectrometry technology (iKnife) along with a deep learning framework for detection of BCC signatures from tissue burns. METHODS: Resected surgical specimen were collected and inspected by a pathologist. With their guidance, data were collected by burning regions of the specimen labeled as BCC or normal, with the iKnife. Data included 190 scans of which 127 were normal and 63 were BCC. A data augmentation approach was proposed by modifying the location and intensity of the peaks of the original spectra, through noise addition in the time and frequency domains. A symmetric autoencoder was built by simultaneously optimizing the spectral reconstruction error and the classification accuracy. Using t-SNE, the latent space was visualized. RESULTS: The autoencoder achieved an accuracy (standard deviation) of 96.62 (1.35%) when classifying BCC and normal scans, a statistically significant improvement over the baseline state-of-the-art approach used in the literature. The t-SNE plot of the latent space distinctly showed the separability between BCC and normal data, not visible with the original data. Augmented data resulted in significant improvements to the classification accuracy of the baseline model. CONCLUSION: We demonstrate the utility of a deep learning framework applied to mass spectrometry data for surgical margin detection. We apply the proposed framework to an application with light surgical overhead and high incidence, the removal of BCC. The learnt models can accurately separate BCC from normal tissue.

Navigated tissue characterization during skin cancer surgery.

PURPOSE: Basal cell carcinoma (BCC) is the most commonly diagnosed skin cancer and is treated by surgical resection. Incomplete tumor removal requires surgical revision, leading to significant healthcare costs and impaired cosmesis. We investigated the clinical feasibility of a surgical navigation system for BCC surgery, based on molecular tissue characterization using rapid evaporative ionization mass spectrometry (REIMS). METHODS: REIMS enables direct tissue characterization by analysis of cell-specific molecules present within surgical smoke, produced during electrocautery tissue resection. A tissue characterization model was built by acquiring REIMS spectra of BCC, healthy skin and fat from ex vivo skin cancer specimens. This model was used for tissue characterization during navigated skin cancer surgery. Navigation was enabled by optical tracking and real-time visualization of the cautery relative to a contoured resection volume. The surgical smoke was aspirated into a mass spectrometer and directly analyzed with REIMS. Classified BCC was annotated at the real-time position of the cautery. Feasibility of the navigation system, and tissue classification accuracy for ex vivo and intraoperative surgery were evaluated. RESULTS: Fifty-four fresh excision specimens were used to build the ex vivo model of BCC, normal skin and fat, with 92% accuracy. While 3 surgeries were successfully navigated without breach of sterility, the intraoperative performance of the ex vivo model was low (< 50%). Hypotheses are: (1) the model was trained on heterogeneous mass spectra that did not originate from a single tissue type, (2) during surgery mixed tissue types were resected and thus presented to the model, and (3) the mass spectra were not validated by pathology. CONCLUSION: REIMS-navigated skin cancer surgery has the potential to detect and localize remaining tumor intraoperatively. Future work will be focused on improving our model by using a precise pencil cautery tip for burning localized tissue types, and having pathology-validated mass spectra.

Real-time wireless tumor tracking during breast conserving surgery.

PURPOSE: To evaluate a novel surgical navigation system for breast conserving surgery (BCS), based on real-time tumor tracking using the Calypso[Formula: see text] 4D Localization System (Varian Medical Systems Inc., USA). Navigation-guided breast conserving surgery (Nav-BCS) was compared to conventional iodine seed-guided BCS ([Formula: see text]I-BCS). METHODS: Two breast phantom types were produced, containing spherical and complex tumors in which wireless transponders (Nav-BCS) or a iodine seed ([Formula: see text]I-BCS) were implanted. For navigation, orthogonal views and 3D volume renders of a CT of the phantom were shown, including a tumor segmentation and a predetermined resection margin. In the same views, a surgical pointer was tracked and visualized. [Formula: see text]I-BCS was performed according to standard protocol. Five surgical breast oncologists first performed a practice session with Nav-BCS, followed by two Nav-BCS and [Formula: see text]I-BCS sessions on spherical and complex tumors. Postoperative CT images of all resection specimens were registered to the preoperative CT. Main outcome measures were the minimum resection margin (in mm) and the excision times. RESULTS: The rate of incomplete tumor resections was 6.7% for Nav-BCS and 20% for [Formula: see text]I-BCS. The minimum resection margins on the spherical tumors were 3.0 ± 1.4 mm for Nav-BCS and 2.5 ± 1.6 mm for [Formula: see text]I-BCS (p = 0.63). For the complex tumors, these were 2.2 ± 1.1 mm (Nav-BCS) and 0.9 ± 2.4 mm ([Formula: see text]I-BCS) (p = 0.32). Mean excision times on spherical and complex tumors were 9.5 ±  2.7 min and 9.4 ± 2.6 min (Nav-BCS), compared to 5.8 ± 2.2  min and 4.7 ± 3.4 min ([Formula: see text]I-BCS, both (p < 0.05). CONCLUSIONS: The presented surgical navigation system improved the intra-operative awareness about tumor position and orientation, with the potential to improve surgical outcomes for non-palpable breast tumors. Results are positive, and participating surgeons were enthusiastic, but extended surgical experience on real breast tissue is required.

The feasibility of manual parameter tuning for deformable breast MR image registration from a multi-objective optimization perspective.

Deformable image registration is typically formulated as an optimization problem involving a linearly weighted combination of terms that correspond to objectives of interest (e.g. similarity, deformation magnitude). The weights, along with multiple other parameters, need to be manually tuned for each application, a task currently addressed mainly via trial-and-error approaches. Such approaches can only be successful if there is a sensible interplay between parameters, objectives, and desired registration outcome. This, however, is not well established. To study this interplay, we use multi-objective optimization, where multiple solutions exist that represent the optimal trade-offs between the objectives, forming a so-called Pareto front. Here, we focus on weight tuning. To study the space a user has to navigate during manual weight tuning, we randomly sample multiple linear combinations. To understand how these combinations relate to desirability of registration outcome, we associate with each outcome a mean target registration error (TRE) based on expert-defined anatomical landmarks. Further, we employ a multi-objective evolutionary algorithm that optimizes the weight combinations, yielding a Pareto front of solutions, which can be directly navigated by the user. To study how the complexity of manual weight tuning changes depending on the registration problem, we consider an easy problem, prone-to-prone breast MR image registration, and a hard problem, prone-to-supine breast MR image registration. Lastly, we investigate how guidance information as an additional objective influences the prone-to-supine registration outcome. Results show that the interplay between weights, objectives, and registration outcome makes manual weight tuning feasible for the prone-to-prone problem, but very challenging for the harder prone-to-supine problem. Here, patient-specific, multi-objective weight optimization is needed, obtaining a mean TRE of 13.6 mm without guidance information reduced to 7.3 mm with guidance information, but also providing a Pareto front that exhibits an intuitively sensible interplay between weights, objectives, and registration outcome, allowing outcome selection.

Supine Breast MRI Using Respiratory Triggering.

RATIONALE AND OBJECTIVES: This study aims to evaluate if navigator-echo respiratory-triggered magnetic resonance acquisition can acquire supine high-quality breast magnetic resonance imaging (MRI). MATERIALS AND METHODS: Supine respiratory-triggered magnetic resonance imaging (Trig-MRI) was compared to supine non-Trig-MRI to evaluate breathing-induced motion artifacts (group 1), and to conventional prone non-Trig-MRI (group 2, 16-channel breast coil), all at 3T. A 32-channel thorax coil was placed on top of a cover to prevent breast deformation. Ten volunteers were scanned in each group, including one patient. The acquisition time was recorded. Image quality was compared by visual examination and by calculation of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and image sharpness (IS). RESULTS: Scan time increased from 56.5 seconds (non-Trig-MRI) to an average of 306 seconds with supine Trig-MRI (range: 120-540 seconds). In group 1, the median values (interquartile range) of SNR, CNR, and IS improved from 11.5 (6.0), 7.3 (3.1), and 0.23 (0.2) cm on supine non-Trig-MRI to 38.1 (29.1), 32.8 (29.7), and 0.12 (0) cm (all P < 0.01) on supine Trig-MRI. All qualitative image parameters in group 1 improved on supine Trig-MRI (all P < 0.01). In group 2, SNR and CNR improved from 14.7 (6.8) and 12.6 (5.6) on prone non-Trig-MRI to 36.2 (12.2) and 32.7 (12.1) (both P < 0.01) on supine Trig-MRI. IS was similar: 0.10 (0) cm vs 0.11 (0) cm (P = 0.88). CONCLUSIONS: Acquisition of high-quality supine breast MRI is possible when respiratory triggering is applied, in a similar setup as during subsequent treatment. Image quality improved when compared to supine non-triggered breast MRI and prone breast MRI, but at the cost of increased acquisition time.

Automatic detection of CT perfusion datasets unsuitable for analysis due to head movement of acute ischemic stroke patients.

Head movement during brain Computed Tomography Perfusion (CTP) can deteriorate perfusion analysis quality in acute ischemic stroke patients. We developed a method for automatic detection of CTP datasets with excessive head movement, based on 3D image-registration of CTP, with non-contrast CT providing transformation parameters. For parameter values exceeding predefined thresholds, the dataset was classified as 'severely moved'. Threshold values were determined by digital CTP phantom experiments. The automated selection was compared to manual screening by 2 experienced radiologists for 114 brain CTP datasets. Based on receiver operator characteristics, optimal thresholds were found of respectively 1.0°, 2.8° and 6.9° for pitch, roll and yaw, and 2.8 mm for z-axis translation. The proposed method had a sensitivity of 91.4% and a specificity of 82.3%. This method allows accurate automated detection of brain CTP datasets that are unsuitable for perfusion analysis.