scholarly journals Challenge for Diagnostic Assessment of Deep Learning Algorithm for Metastases Classification in Sentinel Lymph Nodes on Frozen Tissue Section Digital Slides in Women with Breast Cancer

2020 ◽  
Author(s):  
Young-Gon Kim ◽  
In Hye Song ◽  
Hyunna Lee ◽  
Dong Hyun Yang ◽  
Namkug Kim ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Deep Learning ◽  
Lymph Nodes ◽  
Learning Algorithm ◽  
Medical Center ◽  
Characteristic Curve ◽  
Lobular Carcinoma ◽  
Sentinel Lymph Nodes ◽  
Neoadjuvant Systemic Therapy ◽  
Frozen Tissue

Abstract Assessing the status of metastasis in sentinel lymph nodes (SLNs) by pathologists is an essential task for the accurate staging of breast cancer. However, histopathological evaluation of sentinel lymph nodes by a pathologist is not easy and is a tedious and time-consuming task. The purpose of this study is to review a challenge competition (HeLP 2018) to develop automated solutions for the classification of metastases in hematoxylin and eosin–stained frozen tissue sections of SLNs in breast cancer patients. A total of 297 digital slides were obtained from frozen SLN sections, which include post–neoadjuvant cases (n = 144, 48.5%) in Asan Medical Center, South Korea. The slides were divided into training, development, and validation sets. All of the imaging datasets have been manually segmented by expert pathologists. A total of 10 participants were allowed to use the Kakao challenge platform for six weeks with two P40 GPUs. The algorithms were assessed in terms of the AUC (area under receiver operating characteristic curve). The top three teams showed 0.986, 0.985, and 0.945 AUCs for the development set and 0.805, 0.776, and 0.765 AUCs for the validation set. Micrometastatic tumors, neoadjuvant systemic therapy, invasive lobular carcinoma, and histologic grade 3 were associated with lower diagnostic accuracy. In a challenge competition, accurate deep learning algorithms have been developed, which can be helpful in making frozen diagnosis of intraoperative sentinel lymph node biopsy. Whether this approach has clinical utility will require evaluation in a clinical setting.

scholarly journals Challenge for Diagnostic Assessment of Deep Learning Algorithm for Metastases Classification in Sentinel Lymph Nodes on Frozen Tissue Section Digital Slides in Women with Breast Cancer

2020 ◽  
Author(s):  
Young-Gon Kim ◽  
In Hye Song ◽  
Hyunna Lee ◽  
Dong Hyun Yang ◽  
Namkug Kim ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Deep Learning ◽  
Lymph Nodes ◽  
Tissue Section ◽  
Learning Algorithm ◽  
Sentinel Lymph Nodes ◽  
Diagnostic Assessment ◽  
Deep Learning Algorithm ◽  
Frozen Tissue

Abstract The authors have withdrawn this preprint due to author disagreement.

Frozen Sections of Sentinel Lymph Nodes From Breast Cancer Patients Who Undergo Neoadjuvant Therapy Are Accurately Diagnosed Using Telepathology

2019 ◽  
Vol 152 (Supplement_1) ◽  
pp. S122-S123
Author(s):  
Valarie McMurtry ◽  
Rachel Factor
Keyword(s):  
Breast Cancer ◽  
Lymph Nodes ◽  
Neoadjuvant Therapy ◽  
Cancer Patients ◽  
Treatment Effect ◽  
Frozen Section ◽  
Lobular Carcinoma ◽  
Sentinel Lymph Nodes ◽  
Breast Cancer Patients ◽  
Frozen Sections

Abstract Objectives Telepathology enables histologic diagnosis to be made from a scanned slide visualized on a computer. Frozen sections (FSs) can be performed at remote locations and read by a pathologist at a central site. At our institution, qualifying breast cancer patients are enrolled in clinical trials that require FS on sentinel lymph nodes (SLNs) after neoadjuvant therapy (NAT), including chemotherapy or endocrine therapy. In this setting, histology is complicated by treatment effect and biopsy site changes. Others have reported good accuracy of FSs on SLNs after NAT. We investigated whether pathologists are accurate in diagnosing SLN FSs for such cases while using telepathology. To our knowledge, this has not been reported previously. Methods SLNs were entirely submitted and serially sectioned (2-mm thickness). At least two levels were cut. All FSs were submitted for formalin-fixed, paraffin-embedded permanent sections. A pathology assistant at the remote location prepared the FSs and scanned slides using the VisionTek M6 digital microscope ecosystem (East Dundee, IL). Cases were interpreted by board-certified pathologists who completed training on the VisionTek system. For this study, diagnoses from FSs and permanents were compared. Results Forty-five SLNs from 19 breast neoadjuvant cases were read by VisionTek from March 2017 to January 2019. Forty-three cases (96%) called negative by FSs were confirmed negative (on permanents). One hundred percent called positive by FSs were positive. Four SLN called atypical on FSs were positive. Three of these were neoadjuvant endocrine cases for invasive lobular carcinoma. Two cases called atypical were negative. One of these, called atypical/suspicious, resulted in axillary dissection. This case was reviewed by three pathologists at the time of surgery. It had abundant treatment effect, mimicking carcinoma. Conclusion While pitfalls exist, overall, the diagnostic accuracy of frozen section analysis by telepathology of SLNs from breast cases after neoadjuvant therapy is high.

Prospective Multicenter Comparison of Models to Predict Four or More Involved Axillary Lymph Nodes in Patients With Breast Cancer With One to Three Metastatic Sentinel Lymph Nodes

2009 ◽  
Vol 27 (34) ◽  
pp. 5707-5712 ◽  
Author(s):  
Gabrielle Werkoff ◽  
Eric Lambaudie ◽  
Eric Fondrinier ◽  
Jean Levêque ◽  
Fréderic Marchal ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Lymph Nodes ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
False Negative ◽  
Sentinel Lymph Nodes ◽  
Axillary Lymph Nodes ◽  
Axillary Lymph ◽  
Prospective Multicenter Study ◽  
Bootstrap Estimate

Purpose Three models have been developed to predict four or more involved axillary lymph nodes (ALNs) in patients with breast cancer with one to three involved sentinel lymph nodes (SLNs). Two scores were developed by Chagpar et al (Louisville scores excluding or including method of detection), and a nomogram was developed by Katz et al. The purpose of our investigation was to compare these models in a prospective, multicenter study. Patients and Methods Our study involved a cohort of 536 patients having one to three involved SLNs who underwent ALN dissection. We evaluated the area under the receiver operating characteristic curve (AUC), calibration (for the Katz nomogram only), false-negative (FN) rate, and clinical utility of the three models. Results were compared with the optimal logistic regression (OLR) model that was developed from the validation cohort. Results Among the 536 patients, 57 patients (10.6%) had ≥ four involved ALNs. The AUC for the Katz nomogram was 0.84 (95% CI, 0.81 to 0.86). The Louisville score excluding method of detection was 0.75 (95% CI, 0.72 to 0.78). The Louisville score including method of detection was 0.77 (95% CI, 0.74 to 0.79). The FN rates were 2.5% (eight of 321 patients), 1.8% (two of 109 patients), and 0% (zero of 27 patients) for the Katz nomogram and the Louisville scores excluding and including method of detection, respectively. The Katz nomogram was well calibrated. Optimism-corrected bootstrap estimate AUC of the OLR model was 0.86. Using this result as a reasonable target for an external model, the performance of the Katz nomogram was remarkable. Conclusion We validated the three models for their use in clinical practice. The Katz nomogram outperformed the two other models.

Touch Preparations for the Intraoperative Evaluation of Sentinel Lymph Nodes After Neoadjuvant Therapy Have High False-Negative Rates in Patients With Breast Cancer

2014 ◽  
Vol 138 (6) ◽  
pp. 814-818 ◽  
Author(s):  
Robin M. Elliott ◽  
Robert R. Shenk ◽  
Cheryl L. Thompson ◽  
Hannah L. Gilmore
Keyword(s):  
Breast Cancer ◽  
Neoadjuvant Chemotherapy ◽  
Lymph Nodes ◽  
Neoadjuvant Therapy ◽  
Lobular Carcinoma ◽  
False Negative ◽  
Metastatic Breast ◽  
Sentinel Lymph Nodes ◽  
Intraoperative Evaluation ◽  
Touch Preparation

Context.— The use of a touch preparation for intraoperative sentinel lymph node diagnosis has become a preferred method of many pathologists because of its reported high sensitivity and rapid turnaround time. However, after neoadjuvant chemotherapy many lymph nodes have significant treatment-related changes that may affect the diagnostic accuracy of the intraoperative evaluation. Objective.— To determine the accuracy of touch preparation for the intraoperative diagnosis of metastatic breast carcinoma in the neoadjuvant setting. Design.— We reviewed retrospectively the results of intraoperative evaluations for 148 different sentinel lymph nodes from 63 patients who had undergone neoadjuvant chemotherapy for invasive breast cancer at our institution. The intraoperative touch preparation results were compared with the final pathology reports in conjunction with relevant clinical data. Results.— Use of touch preparation for the evaluation of sentinel lymph nodes intraoperatively after neoadjuvant therapy was associated with a low sensitivity of 38.6% (95% confidence interval [CI], 24.4–54.5) but high specificity of 100% (95% CI, 96.5–100). There was no difference in sensitivity rates between cytopathologists and noncytopathologists in this cohort (P = .40). Patients with invasive lobular carcinoma and those who had a clinically positive axilla before the initiation of neoadjuvant therapy were the most likely to have a false-negative result at surgery. Conclusions.— Intraoperative touch preparations should not be used alone for the evaluation of sentinel lymph nodes in the setting of neoadjuvant therapy for breast cancer because of low overall sensitivity.

scholarly journals Metastatic status of sentinel lymph nodes in breast cancer determined with photoacoustic microscopy via dual-targeting nanoparticles

2020 ◽  
Vol 9 (1) ◽  
Author(s):  
Yanfeng Dai ◽  
Xiang Yu ◽  
Jianshuang Wei ◽  
Fanxin Zeng ◽  
Yiran Li ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Lymph Nodes ◽  
Scavenger Receptor ◽  
Intradermal Injection ◽  
Sentinel Lymph Nodes ◽  
Breast Cancer Surgery ◽  
Photoacoustic Microscopy ◽  
Dual Targeting ◽  
Scavenger Receptor Class ◽  
Distinct Distribution

Abstract Detection of sentinel lymph nodes (SLNs) is critical to guide the treatment of breast cancer. However, distinguishing metastatic SLNs from normal and inflamed lymph nodes (LNs) during surgical resection remains a challenge. Here, we report a CD44 and scavenger receptor class B1 dual-targeting hyaluronic acid nanoparticle (5K-HA-HPPS) loaded with the near-infra-red fluorescent dye DiR-BOA for SLN imaging in breast cancer. The small sized (~40 nm) self-assembled 5K-HA-HPPSs accumulated rapidly in the SLNs after intradermal injection. Compared with normal popliteal LNs (N-LN), there were ~3.2-fold and ~2.4-fold increases in fluorescence intensity in tumour metastatic SLNs (T-MLN) and inflamed LNs (Inf-LN), respectively, 6 h after nanoparticle inoculation. More importantly, photoacoustic microscopy (PAM) of 5K-HA-HPPS showed a significantly distinct distribution in T-MLN compared with N-LN and Inf-LN. Signals were mainly distributed at the centre of T-MLN but at the periphery of N-LN and Inf-LN. The ratio of PA intensity (R) at the centre of the LNs compared with that at the periphery was 5.93 ± 0.75 for T-MLNs of the 5K-HA-HPPS group, which was much higher than that for the Inf-LNs (R = 0.2 ± 0.07) and N-LNs (R = 0.45 ± 0.09). These results suggest that 5K-HA-HPPS injection combined with PAM provides a powerful tool for distinguishing metastatic SLNs from pLNs and inflamed LNs, thus guiding the removal of SLNs during breast cancer surgery.

scholarly journals Development and Verification of a Deep Learning Algorithm to Evaluate Small-Bowel Preparation Quality

Diagnostics ◽  
2021 ◽  
Vol 11 (6) ◽  
pp. 1127
Author(s):  
Ji Hyung Nam ◽  
Dong Jun Oh ◽  
Sumin Lee ◽  
Hyun Joo Song ◽  
Yun Jeong Lim
Keyword(s):  
Deep Learning ◽  
Small Bowel ◽  
Scoring System ◽  
Operating Characteristic ◽  
Clinical Evidence ◽  
Learning Algorithm ◽  
Characteristic Curve ◽  
External Validation ◽  
Test Results ◽  
Deep Learning Algorithm

Capsule endoscopy (CE) quality control requires an objective scoring system to evaluate the preparation of the small bowel (SB). We propose a deep learning algorithm to calculate SB cleansing scores and verify the algorithm’s performance. A 5-point scoring system based on clarity of mucosal visualization was used to develop the deep learning algorithm (400,000 frames; 280,000 for training and 120,000 for testing). External validation was performed using additional CE cases (n = 50), and average cleansing scores (1.0 to 5.0) calculated using the algorithm were compared to clinical grades (A to C) assigned by clinicians. Test results obtained using 120,000 frames exhibited 93% accuracy. The separate CE case exhibited substantial agreement between the deep learning algorithm scores and clinicians’ assessments (Cohen’s kappa: 0.672). In the external validation, the cleansing score decreased with worsening clinical grade (scores of 3.9, 3.2, and 2.5 for grades A, B, and C, respectively, p < 0.001). Receiver operating characteristic curve analysis revealed that a cleansing score cut-off of 2.95 indicated clinically adequate preparation. This algorithm provides an objective and automated cleansing score for evaluating SB preparation for CE. The results of this study will serve as clinical evidence supporting the practical use of deep learning algorithms for evaluating SB preparation quality.

Sign in / Sign up

Export Citation Format

Share Document