Separating Surface Reflectance from Volume Reflectance in Medical Hyperspectral Imaging.

Hyperspectral imaging has shown great promise for diagnostic applications, particularly in cancer surgery. However, non-bulk tissue-related spectral variations complicate the data analysis. Common techniques, such as standard normal variate normalization, often lead to a loss of amplitude and scattering information. This study investigates a novel approach to address these spectral variations in hyperspectral images of optical phantoms and excised human breast tissue. Our method separates surface and volume reflectance, hypothesizing that spectral variability arises from significant variations in surface reflectance across pixels. An illumination setup was developed to measure samples with a hyperspectral camera from different axial positions but with identical zenith angles. This configuration, combined with a novel data analysis approach, allows for the estimation and separation of surface reflectance for each direction and volume reflectance across all directions. Validated with optical phantoms, our method achieved an 83% reduction in spectral variability. Its functionality was further demonstrated in excised human breast tissue. Our method effectively addresses variations caused by surface reflectance or glare while conserving surface reflectance information, which may enhance sample analysis and evaluation. It benefits samples with unknown refractive index spectra and can be easily adapted and applied across a wide range of fields where hyperspectral imaging is used.

Margin assessment during breast conserving surgery using diffuse reflectance spectroscopy.

Significance: During breast-conserving surgeries, it is essential to evaluate the resection margins (edges of breast specimen) to determine whether the tumor has been removed completely. In current surgical practice, there are no methods available to aid in accurate real-time margin evaluation. Aim: In this study, we investigated the diagnostic accuracy of diffuse reflectance spectroscopy (DRS) combined with tissue classification models in discriminating tumorous tissue from healthy tissue up to 2 mm in depth on the actual resection margin of in vivo breast tissue. Approach: We collected an extensive dataset of DRS measurements on ex vivo breast tissue and in vivo breast tissue, which we used to develop different classification models for tissue classification. Next, these models were used in vivo to evaluate the performance of DRS for tissue discrimination during breast conserving surgery. We investigated which training strategy yielded optimum results for the classification model with the highest performance. Results: We achieved a Matthews correlation coefficient of 0.76, a sensitivity of 96.7% (95% CI 95.6% to 98.2%), a specificity of 90.6% (95% CI 86.3% to 97.9%) and an area under the curve of 0.98 by training the optimum model on a combination of ex vivo and in vivo DRS data. Conclusions: DRS allows real-time margin assessment with a high sensitivity and specificity during breast-conserving surgeries.

Spatial and Spectral Reconstruction of Breast Lumpectomy Hyperspectral Images.

(1) Background: Hyperspectral imaging has emerged as a promising margin assessment technique for breast-conserving surgery. However, to be implicated intraoperatively, it should be both fast and capable of yielding high-quality images to provide accurate guidance and decision-making throughout the surgery. As there exists a trade-off between image quality and data acquisition time, higher resolution images come at the cost of longer acquisition times and vice versa. (2) Methods: Therefore, in this study, we introduce a deep learning spatial-spectral reconstruction framework to obtain a high-resolution hyperspectral image from a low-resolution hyperspectral image combined with a high-resolution RGB image as input. (3) Results: Using the framework, we demonstrate the ability to perform a fast data acquisition during surgery while maintaining a high image quality, even in complex scenarios where challenges arise, such as blur due to motion artifacts, dead pixels on the camera sensor, noise from the sensor's reduced sensitivity at spectral extremities, and specular reflections caused by smooth surface areas of the tissue. (4) Conclusion: This gives the opportunity to facilitate an accurate margin assessment through intraoperative hyperspectral imaging.

Tumor Segmentation in Colorectal Ultrasound Images Using an Ensemble Transfer Learning Model: Towards Intra-Operative Margin Assessment.

Tumor boundary identification during colorectal cancer surgery can be challenging, and incomplete tumor removal occurs in approximately 10% of the patients operated for advanced rectal cancer. In this paper, a deep learning framework for automatic tumor segmentation in colorectal ultrasound images was developed, to provide real-time guidance on resection margins using intra-operative ultrasound. A colorectal ultrasound dataset was acquired consisting of 179 images from 74 patients, with ground truth tumor annotations based on histopathology results. To address data scarcity, transfer learning techniques were used to optimize models pre-trained on breast ultrasound data for colorectal ultrasound data. A new custom gradient-based loss function (GWDice) was developed, which emphasizes the clinically relevant top margin of the tumor while training the networks. Lastly, ensemble learning methods were applied to combine tumor segmentation predictions of multiple individual models and further improve the overall tumor segmentation performance. Transfer learning outperformed training from scratch, with an average Dice coefficient over all individual networks of 0.78 compared to 0.68. The new GWDice loss function clearly decreased the average tumor margin prediction error from 1.08 mm to 0.92 mm, without compromising the segmentation of the overall tumor contour. Ensemble learning further improved the Dice coefficient to 0.84 and the tumor margin prediction error to 0.67 mm. Using transfer and ensemble learning strategies, good tumor segmentation performance was achieved despite the relatively small dataset. The developed US segmentation model may contribute to more accurate colorectal tumor resections by providing real-time intra-operative feedback on tumor margins.

Diffuse reflectance spectroscopy for accurate margin assessment in breast-conserving surgeries: importance of an optimal number of fibers.

During breast-conserving surgeries, it remains challenging to accomplish adequate surgical margins. We investigated different numbers of fibers for fiber-optic diffuse reflectance spectroscopy to differentiate tumorous breast tissue from healthy tissue ex vivo up to 2 mm from the margin. Using a machine-learning classification model, the optimal performance was obtained using at least three emitting fibers (Matthew's correlation coefficient (MCC) of 0.73), which was significantly higher compared to the performance of using a single-emitting fiber (MCC of 0.48). The percentage of correctly classified tumor locations varied from 75% to 100% depending on the tumor percentage, the tumor-margin distance and the number of fibers.

Tissue Classification of Breast Cancer by Hyperspectral Unmixing.

(1) Background: Assessing the resection margins during breast-conserving surgery is an important clinical need to minimize the risk of recurrent breast cancer. However, currently there is no technique that can provide real-time feedback to aid surgeons in the margin assessment. Hyperspectral imaging has the potential to overcome this problem. To classify resection margins with this technique, a tissue discrimination model should be developed, which requires a dataset with accurate ground-truth labels. However, establishing such a dataset for resection specimens is difficult. (2) Methods: In this study, we therefore propose a novel approach based on hyperspectral unmixing to determine which pixels within hyperspectral images should be assigned to the ground-truth labels from histopathology. Subsequently, we use this hyperspectral-unmixing-based approach to develop a tissue discrimination model on the presence of tumor tissue within the resection margins of ex vivo breast lumpectomy specimens. (3) Results: In total, 372 measured locations were included on the lumpectomy resection surface of 189 patients. We achieved a sensitivity of 0.94, specificity of 0.85, accuracy of 0.87, Matthew's correlation coefficient of 0.71, and area under the curve of 0.92. (4) Conclusion: Using this hyperspectral-unmixing-based approach, we demonstrated that the measured locations with hyperspectral imaging on the resection surface of lumpectomy specimens could be classified with excellent performance.

Toward Intraoperative Margin Assessment Using a Deep Learning-Based Approach for Automatic Tumor Segmentation in Breast Lumpectomy Ultrasound Images.

There is an unmet clinical need for an accurate, rapid and reliable tool for margin assessment during breast-conserving surgeries. Ultrasound offers the potential for a rapid, reproducible, and non-invasive method to assess margins. However, it is challenged by certain drawbacks, including a low signal-to-noise ratio, artifacts, and the need for experience with the acquirement and interpretation of images. A possible solution might be computer-aided ultrasound evaluation. In this study, we have developed new ensemble approaches for automated breast tumor segmentation. The ensemble approaches to predict positive and close margins (distance from tumor to margin ≤ 2.0 mm) in the ultrasound images were based on 8 pre-trained deep neural networks. The best optimum ensemble approach for segmentation attained a median Dice score of 0.88 on our data set. Furthermore, utilizing the segmentation results we were able to achieve a sensitivity of 96% and a specificity of 76% for predicting a close margin when compared to histology results. The promising results demonstrate the capability of AI-based ultrasound imaging as an intraoperative surgical margin assessment tool during breast-conserving surgery.

Discriminating healthy from tumor tissue in breast lumpectomy specimens using deep learning-based hyperspectral imaging.

Achieving an adequate resection margin during breast-conserving surgery remains challenging due to the lack of intraoperative feedback. Here, we evaluated the use of hyperspectral imaging to discriminate healthy tissue from tumor tissue in lumpectomy specimens. We first used a dataset obtained on tissue slices to develop and evaluate three convolutional neural networks. Second, we fine-tuned the networks with lumpectomy data to predict the tissue percentages of the lumpectomy resection surface. A MCC of 0.92 was achieved on the tissue slices and an RMSE of 9% on the lumpectomy resection surface. This shows the potential of hyperspectral imaging to classify the resection margins of lumpectomy specimens.