Correlation of breast tissue histology and optical signatures to improve margin assessment techniques

Evaluation of an Integrated Spectroscopy and Classification Platform for Point-of-Care Core Needle Biopsy Assessment: Performance Characteristics from Ex Vivo Renal Mass Biopsies

PURPOSE

To evaluate a transmission optical spectroscopy instrument for rapid ex vivo assessment of core needle cancer biopsies (CNBs) at the point of care.

MATERIALS AND METHODS

CNBs from surgically resected renal tumors and nontumor regions were scanned on their sampling trays with a custom spectroscopy instrument. After extracting principal spectral components, machine learning was used to train logistic regression, support vector machines, and random decision forest (RF) classifiers on 80% of randomized and stratified data. The algorithms were evaluated on the remaining 20% of the data set held out during training. Binary classification (tumor/nontumor) was performed based on a decision threshold. Multinomial classification was also performed to differentiate between the subtypes of renal cell carcinoma (RCC) and account for potential confounding effects from fat, blood, and necrotic tissue. Classifiers were compared based on sensitivity, specificity, and positive predictive value (PPV) relative to a histopathologic standard.

RESULTS

A total of 545 CNBs from 102 patients were analyzed, yielding 5,583 spectra after outlier exclusion. At the individual spectra level, the best performing algorithm was RF with sensitivities of 96% and 92% and specificities of 90% and 89%, for the binary and multiclass analyses, respectively. At the full CNB level, RF algorithm also showed the highest sensitivity and specificity (93% and 91%, respectively). For RCC subtypes, the highest sensitivity and PPV were attained for clear cell (93.5%) and chromophobe (98.2%) subtypes, respectively.

CONCLUSIONS

Ex vivo spectroscopy imaging paired with machine learning can accurately characterize renal mass CNB at the time of tissue acquisition.

Histopathology Feature Mining and Association with Hyperspectral Imaging for the Detection of Squamous Neoplasia

Hyperspectral imaging (HSI) is a noninvasive optical modality that holds promise for early detection of tongue lesions. Spectral signatures generated by HSI contain important diagnostic information that can be used to predict the disease status of the examined biological tissue. However, the underlying pathophysiology for the spectral difference between normal and neoplastic tissue is not well understood. Here, we propose to leverage digital pathology and predictive modeling to select the most discriminative features from digitized histological images to differentiate tongue neoplasia from normal tissue, and then correlate these discriminative pathological features with corresponding spectral signatures of the neoplasia. We demonstrated the association between the histological features quantifying the architectural features of neoplasia on a microscopic scale, with the spectral signature of the corresponding tissue measured by HSI on a macroscopic level. This study may provide insight into the pathophysiology underlying the hyperspectral dataset.

Wide-field optical property mapping and structured light imaging of the esophagus with spatial frequency domain imaging

As the incidence of esophageal adenocarcinoma continues to rise, there is a need for improved imaging technologies with contrast to abnormal esophageal tissues. To inform the design of optical technologies that meet this need, we characterize the spatial distribution of the scattering and absorption properties from 471 to 851 nm of eight resected human esophagi tissues using Spatial Frequency Domain Imaging. Histopathology was used to categorize tissue types, including normal, inflammation, fibrotic, ulceration, Barrett's Esophagus and squamous cell carcinoma. Average absorption and reduced scattering coefficients of normal tissues were 0.211 ± 0.051 and 1.20 ± 0.18 mm-1 , respectively at 471 nm, and both values decreased monotonically with increasing wavelength. Fibrotic tissue exhibited at least 68% larger scattering signal across all wavelengths, while squamous cell carcinoma exhibited a 36% decrease in scattering at 471 nm. We additionally image the esophagus with high spatial frequencies up to 0.5 mm-1 and show strong reflectance contrast to tissue treated with radiation. Lastly, we observe that esophageal absorption and scattering values change by an average of 9.4% and 2.7% respectively over a 30 minute duration post-resection. These results may guide system design for the diagnosis, prevention and monitoring of esophageal pathologies.

Towards the use of diffuse reflectance spectroscopy for real-time in vivo detection of breast cancer during surgery

Background

Breast cancer surgeons struggle with differentiating healthy tissue from cancer at the resection margin during surgery. We report on the feasibility of using diffuse reflectance spectroscopy (DRS) for real-time in vivo tissue characterization.

Methods

Evaluating feasibility of the technology requires a setting in which measurements, imaging and pathology have the best possible correlation. For this purpose an optical biopsy needle was used that had integrated optical fibers at the tip of the needle. This approach enabled the best possible correlation between optical measurement volume and tissue histology. With this optical biopsy needle we acquired real-time DRS data of normal tissue and tumor tissue in 27 patients that underwent an ultrasound guided breast biopsy procedure. Five additional patients were measured in continuous mode in which we obtained DRS measurements along the entire biopsy needle trajectory. We developed and compared three different support vector machine based classification models to classify the DRS measurements.

Results

With DRS malignant tissue could be discriminated from healthy tissue. The classification model that was based on eight selected wavelengths had the highest accuracy and Matthews Correlation Coefficient (MCC) of 0.93 and 0.87, respectively. In three patients that were measured in continuous mode and had malignant tissue in their biopsy specimen, a clear transition was seen in the classified DRS measurements going from healthy tissue to tumor tissue. This transition was not seen in the other two continuously measured patients that had benign tissue in their biopsy specimen.

Conclusions

It was concluded that DRS is feasible for integration in a surgical tool that could assist the breast surgeon in detecting positive resection margins during breast surgery.

Trail registration NIH US National Library of Medicine–clinicaltrails.gov, NCT01730365. Registered: 10/04/2012 https://clinicaltrials.gov/ct2/show/study/NCT01730365

Light scattering measured with spatial frequency domain imaging can predict stromal versus epithelial proportions in surgically resected breast tissue

This study aims to determine if light scatter parameters measured with spatial frequency domain imaging (SFDI) can accurately predict stromal, epithelial, and adipose fractions in freshly resected, unstained human breast specimens. An explicit model was developed to predict stromal, epithelial, and adipose fractions as a function of light scattering parameters, which was validated against a quantitative analysis of digitized histology slides for N  =  31 specimens using leave-one-out cross-fold validation. Specimen mean stromal, epithelial, and adipose volume fractions predicted from light scattering parameters strongly correlated with those calculated from digitized histology slides (r  =  0.90, 0.77, and 0.91, respectively, p-value <1  ×  10  -  6). Additionally, the ratio of predicted epithelium to stroma classified malignant specimens with a sensitivity and specificity of 90% and 81%, respectively, and also classified all pixels in malignant lesions with 63% and 79%, at a threshold of 1. All specimens and pixels were classified as malignant, benign, or fat with 84% and 75% accuracy, respectively. These findings demonstrate how light scattering parameters acquired with SFDI can be used to accurately predict and spatially map stromal, epithelial, and adipose proportions in fresh unstained, human breast tissue, and suggest that these estimations could provide diagnostic value.

Automated detection of breast cancer in resected specimens with fluorescence lifetime imaging

Re-excision rates for breast cancer lumpectomy procedures are currently nearly 25% due to surgeons relying on inaccurate or incomplete methods of evaluating specimen margins. The objective of this study was to determine if cancer could be automatically detected in breast specimens from mastectomy and lumpectomy procedures by a classification algorithm that incorporated parameters derived from fluorescence lifetime imaging (FLIm). This study generated a database of co-registered histologic sections and FLIm data from breast cancer specimens (N  =  20) and a support vector machine (SVM) classification algorithm able to automatically detect cancerous, fibrous, and adipose breast tissue. Classification accuracies were greater than 97% for automated detection of cancerous, fibrous, and adipose tissue from breast cancer specimens. The classification worked equally well for specimens scanned by hand or with a mechanical stage, demonstrating that the system could be used during surgery or on excised specimens. The ability of this technique to simply discriminate between cancerous and normal breast tissue, in particular to distinguish fibrous breast tissue from tumor, which is notoriously challenging for optical techniques, leads to the conclusion that FLIm has great potential to assess breast cancer margins. Identification of positive margins before waiting for complete histologic analysis could significantly reduce breast cancer re-excision rates.

Label-free optical imaging technologies for rapid translation and use during intraoperative surgical and tumor margin assessment

The biannual International Conference on Biophotonics was recently held on April 30 to May 1, 2017, in Fremantle, Western Australia. This continuing conference series brought together key opinion leaders in biophotonics to present their latest results and, importantly, to participate in discussions on the future of the field and what opportunities exist when we collectively work together for using biophotonics for biological discovery and medical applications. One session in this conference, entitled "Tumor Margin Identification: Critiquing Technologies," challenged invited speakers and attendees to review and critique representative label-free optical imaging technologies and their application for intraoperative assessment and guidance in surgical oncology. We are pleased to share a summary in this outlook paper, with the intent to motivate more research inquiry and investigations, to challenge these and other optical imaging modalities to evaluate and improve performance, to spur translation and adoption, and ultimately, to improve the care and outcomes of patients.