Design and evaluation of an AR-based thermal imaging system for planning reconstructive surgeries.

INTRODUCTION: Thermal imaging can be used for the non-invasive detection of blood vessels of the skin. However, mapping the results to the patient currently lacks user-friendliness. Augmented reality may provide a useful tool to superimpose thermal information on the patient. METHODS: A system to support planning in reconstructive surgery using a thermal camera was designed. The obtained information was superimposed on the physical object using a Microsoft HoloLens. An RGB, depth, and thermal camera were combined to capture a scene of different modalities and reconstruct a virtual scene in real time. To register the different cameras and the AR device, an active calibration target was developed and evaluated. A Vuforia marker was used to register the hologram in the virtual space. The accuracy of the projected hologram was evaluated in a laboratory setting with participants by measuring the error between the physical object and the hologram. RESULTS: The AR-based system was evaluated by 21 participants in a laboratory setting. The mean projection error is 10.3 ± 9.4 mm. The system is able to stream a three-dimensional scene with augmented thermal information in real time at 5 frames per second. The active calibration target can be used independently of the environment. CONCLUSION: The calibration target provides an easy-to-use method for the registration of cameras capturing the visible to long-infrared spectral range. The inside-out tracking of the HoloLens in combination with a Vuforia marker is not accurate enough for the intended clinical use.

Switchable LED-based laparoscopic multispectral system for rapid high-resolution perfusion imaging.

Significance: Multispectral imaging (MSI) is an approach for real-time, quantitative, and non-invasive tissue perfusion measurements. Current laparoscopic systems based on mosaic sensors or filter wheels lack high spatial resolution or acceptable frame rates. Aim: To develop a laparoscopic system for MSI-based color video and tissue perfusion imaging during gastrointestinal surgery without compromising spatial or temporal resolution. Approach: The system was built with 14 switchable light-emitting diodes in the visible and near-infrared spectral range, a 4K image sensor, and a 10 mm laparoscope. Illumination patterns were created for tissue oxygenation and hemoglobin content monitoring. The system was calibrated to a clinically approved laparoscopic hyperspectral system using linear regression models and evaluated in an occlusion study with 36 volunteers. Results: The root mean squared errors between the MSI and reference system were 0.073 for hemoglobin content, 0.039 for oxygenation in deeper tissue layers, and 0.093 for superficial oxygenation. The spatial resolution at a working distance of 45 mm was 156 µm. The effective frame rate was 20 fps. Conclusions: High-resolution perfusion monitoring was successfully achieved. Hardware optimizations will increase the frame rate. Parameter optimizations through alternative illumination patterns, regression, or assumed tissue models are planned. Intraoperative measurements must confirm the suitability during surgery.

Discrimination of human and animal bloodstains using hyperspectral imaging.

Blood is the most encountered type of biological evidence in violent crimes and contains pertinent information to a forensic investigation. The false presumption that blood encountered at a crime scene is human may not be realised until after costly and sample-consuming tests are performed. To address the question of blood origin, the novel application of visible-near infrared hyperspectral imaging (HSI) is used for the detection and discrimination of human and animal bloodstains. The HSI system used is a portable, non-contact, non-destructive method for the determination of blood origin. A support vector machine (SVM) binary classifier was trained for the discrimination of bloodstains of human (n = 20) and five animal species: pig (n = 20), mouse (n = 16), rat (n = 5), rabbit (n = 5), and cow (n = 20). On an independent test set, the SVM model achieved accuracy, precision, sensitivity, and specificity values of 96, 97, 95, and 96%, respectively. Segmented images of bloodstains aged over a period of two months were produced, allowing for the clear visualisation of the discrimination of human and animal bloodstains. The inclusion of such a system in a forensic investigation workflow not only removes ambiguity surrounding blood origin, but can potentially be used in tandem with HSI bloodstain age determination methods for rapid on-scene forensic analysis.

Impact of Pre- and Post-Processing Steps for Supervised Classification of Colorectal Cancer in Hyperspectral Images.

BACKGROUND: Recent studies have shown that hyperspectral imaging (HSI) combined with neural networks can detect colorectal cancer. Usually, different pre-processing techniques (e.g., wavelength selection and scaling, smoothing, denoising) are analyzed in detail to achieve a well-trained network. The impact of post-processing was studied less. METHODS: We tested the following methods: (1) Two pre-processing techniques (Standardization and Normalization), with (2) Two 3D-CNN models: Inception-based and RemoteSensing (RS)-based, with (3) Two post-processing algorithms based on median filter: one applies a median filter to a raw predictions map, the other applies the filter to the predictions map after adopting a discrimination threshold. These approaches were evaluated on a dataset that contains ex vivo hyperspectral (HS) colorectal cancer records of 56 patients. RESULTS: (1) Inception-based models perform better than RS-based, with the best results being 92% sensitivity and 94% specificity; (2) Inception-based models perform better with Normalization, RS-based with Standardization; (3) Our outcomes show that the post-processing step improves sensitivity and specificity by 6.6% in total. It was also found that both post-processing algorithms have the same effect, and this behavior was explained. CONCLUSION: HSI combined with tissue classification algorithms is a promising diagnostic approach whose performance can be additionally improved by the application of the right combination of pre- and post-processing.

The role of intraoperative hyperspectral imaging (HSI) in colon interposition after esophagectomy.

BACKGROUND: Colon conduit is an alternative approach to reconstructing the alimentary tract after esophagectomy. Hyperspectral imaging (HSI) has been demonstrated to be effective for evaluating the perfusion of gastric conduits, but not colon conduits. This is the first study to describe this new tool addressing image-guided surgery and supporting esophageal surgeons to select the optimal colon segment for the conduit and anastomotic site intraoperatively. PATIENTS AND METHODS: Of 10 patients, eight who underwent reconstruction with a long-segment colon conduit after esophagectomy between 01/05/2018 and 01/04/2022 were included in this study. HSI was recorded at the root and tip of the colon conduit after clamping the middle colic vessels, allowing us to evaluate the perfusion and appropriate part of the colon segment. RESULTS: Anastomotic leak (AL) was detected in only one (12.5%) of all the enrolled patients (n = 8). None of the patients developed conduit necrosis. Only one patient required re-anastomosis on postoperative day 4. No patient needed conduit removal, esophageal diversion, or stent placement. There was a change in the anastomosis site to proximal in two patients intraoperatively. There was no need to change the side of colon conduit intraoperatively in any patient. CONCLUSION: HSI is a promising and novel intraoperative imaging tool to objectively assess the perfusion of the colon conduit. It helps the surgeon to define the best perfused anastomosis site and the side of colon conduit in this type of operation.

In vivo evaluation of a hyperspectral imaging system for minimally invasive surgery (HSI-MIS).

BACKGROUND: Hyperspectral Imaging (HSI) is a reliable and safe imaging method for taking intraoperative perfusion measurements. This is the first study translating intraoperative HSI to an in vivo laparoscopic setting using a CE-certified HSI-system for minimally invasive surgery (HSI-MIS). We aim to compare it to an established HSI-system for open surgery (HSI-Open). METHODS: Intraoperative HSI was done using the HSI-MIS and HSI-Open at the Region of Interest (ROI). 19 patients undergoing gastrointestinal resections were analyzed in this study. The HSI-MIS-acquired images were aligned with those from the HSI-Open, and spectra and parameter images were compared pixel-wise. We calculated the Mean Absolute Error (MAE) for Tissue Oxygen Saturation (StO2), Near-Infrared Perfusion Index (NIR-PI), Tissue Water Index (TWI), and Organ Hemoglobin Index (OHI), as well as the Root Mean Squared Error (RMSE) over the whole spectrum. Our analysis of parameters was optimized using partial least squares (PLS) regression. Two experienced surgeons carried out an additional color-change analysis, comparing the ROI images and deciding whether they provided the same (acceptable) or different visual information (rejected). RESULTS: HSI and subsequent image registration was possible in 19 patients. MAE results for the original calculation were StO2 orig. 17.2% (± 7.7%), NIR-PIorig. 16.0 (± 9.5), TWIorig. 18.1 (± 7.9), OHIorig. 14.4 (± 4.5). For the PLS calculation, they were StO2 PLS 12.6% (± 5.2%), NIR-PIPLS 10.3 (± 6.0), TWIPLS 10.6 (± 5.1), and OHIPLS 11.6 (± 3.0). The RMSE between both systems was 0.14 (± 0.06). In the color-change analysis; both surgeons accepted more images generated using the PLS method. CONCLUSION: Intraoperative HSI-MIS is a new technology and holds great potential for future applications in surgery. Parameter deviations are attributable to technical differences and can be reduced by applying improved calculation methods. This study is an important step toward the clinical implementation of HSI for minimally invasive surgery.

Alternative intraoperative optical imaging modalities for fluorescence angiography in gastrointestinal surgery: spectral imaging and imaging photoplethysmography.

INTRODUCTION: Intraoperative near-infrared fluorescence angiography with indocyanine green (ICG-FA) is a well-established modality in gastrointestinal surgery. Its main drawback is the application of a fluorescent agent with possible side effects for patients. The goal of this review paper is the presentation of alternative, non-invasive optical imaging methods and their comparison with ICG-FA. MATERIAL AND METHODS: The principles of ICG-FA, spectral imaging, imaging photoplethysmography (iPPG), and their applications in gastrointestinal surgery are described based on selected published works. RESULTS: The main applications of the three modalities are the evaluation of tissue perfusion, the identification of risk structures, and tissue segmentation or classification. While the ICG-FA images are mainly evaluated visually, leading to subjective interpretations, quantitative physiological parameters and tissue segmentation are provided in spectral imaging and iPPG. The combination of ICG-FA and spectral imaging is a promising method. CONCLUSIONS: Non-invasive spectral imaging and iPPG have shown promising results in gastrointestinal surgery. They can overcome the main drawbacks of ICG-FA, i.e. the use of contrast agents, the lack of quantitative analysis, repeatability, and a difficult standardization of the acquisition. Further technical improvements and clinical evaluations are necessary to establish them in daily clinical routine.

Spectral Similarity Measures for In Vivo Human Tissue Discrimination Based on Hyperspectral Imaging.

PROBLEM: Similarity measures are widely used as an approved method for spectral discrimination or identification with their applications in different areas of scientific research. Even though a range of works have been presented, only a few showed slightly promising results for human tissue, and these were mostly focused on pathological and non-pathological tissue classification. METHODS: In this work, several spectral similarity measures on hyperspectral (HS) images of in vivo human tissue were evaluated for tissue discrimination purposes. Moreover, we introduced two new hybrid spectral measures, called SID-JM-TAN(SAM) and SID-JM-TAN(SCA). We analyzed spectral signatures obtained from 13 different human tissue types and two different materials (gauze, instruments), collected from HS images of 100 patients during surgeries. RESULTS: The quantitative results showed the reliable performance of the different similarity measures and the proposed hybrid measures for tissue discrimination purposes. The latter produced higher discrimination values, up to 6.7 times more than the classical spectral similarity measures. Moreover, an application of the similarity measures was presented to support the annotations of the HS images. We showed that the automatic checking of tissue-annotated thyroid and colon tissues was successful in 73% and 60% of the total spectra, respectively. The hybrid measures showed the highest performance. Furthermore, the automatic labeling of wrongly annotated tissues was similar for all measures, with an accuracy of up to 90%. CONCLUSION: In future work, the proposed spectral similarity measures will be integrated with tools to support physicians in annotations and tissue labeling of HS images.

Evaluation of Preprocessing Methods on Independent Medical Hyperspectral Databases to Improve Analysis.

Currently, one of the most common causes of death worldwide is cancer. The development of innovative methods to support the early and accurate detection of cancers is required to increase the recovery rate of patients. Several studies have shown that medical Hyperspectral Imaging (HSI) combined with artificial intelligence algorithms is a powerful tool for cancer detection. Various preprocessing methods are commonly applied to hyperspectral data to improve the performance of the algorithms. However, there is currently no standard for these methods, and no studies have compared them so far in the medical field. In this work, we evaluated different combinations of preprocessing steps, including spatial and spectral smoothing, Min-Max scaling, Standard Normal Variate normalization, and a median spatial smoothing technique, with the goal of improving tumor detection in three different HSI databases concerning colorectal, esophagogastric, and brain cancers. Two machine learning and deep learning models were used to perform the pixel-wise classification. The results showed that the choice of preprocessing method affects the performance of tumor identification. The method that showed slightly better results with respect to identifing colorectal tumors was Median Filter preprocessing (0.94 of area under the curve). On the other hand, esophagogastric and brain tumors were more accurately identified using Min-Max scaling preprocessing (0.93 and 0.92 of area under the curve, respectively). However, it is observed that the Median Filter method smooths sharp spectral features, resulting in high variability in the classification performance. Therefore, based on these results, obtained with different databases acquired by different HSI instrumentation, the most relevant preprocessing technique identified in this work is Min-Max scaling.

Hyperspectral imaging detects perfusion and oxygenation differences between stapled and hand-sewn intestinal anastomoses.

Objectives: Hand-sewn and stapled intestinal anastomoses are both daily performed routine procedures by surgeons. Yet, differences in micro perfusion of these two surgical techniques and their impact on surgical outcomes are still insufficiently understood. Only recently, hyperspectral imaging (HSI) has been established as a non-invasive, contact-free, real-time assessment tool for tissue oxygenation and micro-perfusion. Hence, objective of this study was HSI assessment of different intestinal anastomotic techniques and analysis of patients' clinical outcome. Methods: Forty-six consecutive patients with an ileal-ileal anastomoses were included in our study; 21 side-to-side stapled and 25 end-to-end hand-sewn. Based on adsorption and reflectance of the analyzed tissue, chemical color imaging indicates oxygen saturation (StO2), tissue perfusion (near-infrared perfusion index [NIR]), organ hemoglobin index (OHI), and tissue water index (TWI). Results: StO2 as well as NIR of the region of interest (ROI) was significantly higher in stapled anastomoses as compared to hand-sewn ileal-ileal anastomoses (StO2 0.79 (0.74-0.81) vs. 0.66 (0.62-0.70); p<0.001 NIR 0.83 (0.70-0.86) vs. 0.70 (0.63-0.76); p=0.01). In both groups, neither anastomotic leakage nor abdominal septic complications nor patient death did occur. Conclusions: Intraoperative HSI assessment is able to detect significant differences in tissue oxygenation and NIR of hand-sewn and stapled intestinal anastomoses. Long-term clinical consequences resulting from the reduced tissue oxygenation and tissue perfusion in hand-sewn anastomoses need to be evaluated in larger clinical trials, as patients may benefit from further refined surgical techniques.

Comparison of image registration methods for combining laparoscopic video and spectral image data.

Laparoscopic procedures can be assisted by intraoperative modalities, such as quantitative perfusion imaging based on fluorescence or hyperspectral data. If these modalities are not available at video frame rate, fast image registration is needed for the visualization in augmented reality. Three feature-based algorithms and one pre-trained deep homography neural network (DH-NN) were tested for single and multi-homography estimation. Fine-tuning was used to bridge the domain gap of the DH-NN for non-rigid registration of laparoscopic images. The methods were validated on two datasets: an open-source record of 750 manually annotated laparoscopic images, presented in this work, and in-vivo data from a novel laparoscopic hyperspectral imaging system. All feature-based single homography methods outperformed the fine-tuned DH-NN in terms of reprojection error, Structural Similarity Index Measure, and processing time. The feature detector and descriptor ORB1000 enabled video-rate registration of laparoscopic images on standard hardware with submillimeter accuracy.

[Artificial intelligence and hyperspectral imaging for image-guided assistance in minimally invasive surgery]. / Künstliche Intelligenz und hyperspektrale Bildgebung zur bildgestützten Assistenz in der minimal-invasiven Chirurgie.

BACKGROUND: Intraoperative imaging assists surgeons during minimally invasive procedures. Hyperspectral imaging (HSI) is a noninvasive and noncontact optical technique with great diagnostic potential in medicine. The combination with artificial intelligence (AI) approaches to analyze HSI data is called intelligent HSI in this article. OBJECTIVE: What are the medical applications and advantages of intelligent HSI for minimally invasive visceral surgery? MATERIAL AND METHODS: Within various clinical studies HSI data from multiple in vivo tissue types and oncological resections were acquired using an HSI camera system. Different AI algorithms were evaluated for detection and discrimination of organs, risk structures and tumors. RESULTS: In an experimental animal study 20 different organs could be differentiated with high precision (> 95%) using AI. In vivo, the parathyroid glands could be discriminated from surrounding tissue with an F1 score of 47% and sensitivity of 75%, and the bile duct with an F1 score of 79% and sensitivity of 90%. Furthermore, ex vivo tumor tissue could be successfully detected with an area under the receiver operating characteristic (ROC) curve (AUC) larger than 0.91. DISCUSSION: This study demonstrates that intelligent HSI can automatically and accurately detect different tissue types. Despite great progress in the last decade intelligent HSI still has limitations. Thus, accurate AI algorithms that are easier to understand for the user and an extensive standardized and continuously growing database are needed. Further clinical studies should support the various medical applications and lead to the adoption of intelligent HSI in the clinical routine practice.

Video: Clinical evaluation of a laparoscopic hyperspectral imaging system.

BACKGROUND: Hyperspectral imaging (HSI) during surgical procedures is a new method for perfusion quantification and tissue discrimination. Its use has been limited to open surgery due to large camera sizes, missing color video, or long acquisition times. A hand-held, laparoscopic hyperspectral camera has been developed now to overcome those disadvantages and evaluated clinically for the first time. METHODS: In a clinical evaluation study, gastrointestinal resectates of ten cancer patients were investigated using the laparoscopic hyperspectral camera. Reference data from corresponding anatomical regions were acquired with a clinically approved HSI system. An image registration process was executed that allowed for pixel-wise comparisons of spectral data and parameter images (StO2: oxygen saturation of tissue, NIR PI: near-infrared perfusion index, OHI: organ hemoglobin index, TWI: tissue water index) provided by both camera systems. The mean absolute error (MAE) and root mean square error (RMSE) served for the quantitative evaluations. Spearman's rank correlation between factors related to the study design like the time of spectral white balancing and MAE, respectively RMSE, was calculated. RESULTS: The obtained mean MAEs between the TIVITA® Tissue and the laparoscopic hyperspectral system resulted in StO2: 11% ± 7%, NIR PI: 14±3, OHI: 14± 5, and TWI: 10 ± 2. The mean RMSE between both systems was 0.1±0.03 from 500 to 750 nm and 0.15 ±0.06 from 750 to 1000 nm. Spearman's rank correlation coefficients showed no significant correlation between MAE or RMSE and influencing factors related to the study design. CONCLUSION: Qualitatively, parameter images of the laparoscopic system corresponded to those of the system for open surgery. Quantitative deviations were attributed to technical differences rather than the study design. Limitations of the presented study are addressed in current large-scale in vivo trials.

Border Line Definition Using Hyperspectral Imaging in Colorectal Resections.

BACKGROUND: A perfusion deficit is a well-defined and intraoperatively influenceable cause of anastomotic leak (AL). Current intraoperative perfusion assessment methods do not provide objective and quantitative results. In this study, the ability of hyperspectral imaging (HSI) to quantify tissue oxygenation intraoperatively was assessed. METHODS: 115 patients undergoing colorectal resections were included in the final analysis. Before anastomotic formation, the bowel was extracted and the resection line was outlined and imaged using a compact HSI camera, in order to provide instantaneously quantitative perfusion assessment. RESULTS: In 105 patients, a clear demarcation line was visible with HSI one minute after marginal artery transection, reaching a plateau after 3 min. In 58 (55.2%) patients, the clinically determined transection line matched with HSI. In 23 (21.9%) patients, the clinically established resection margin was entirely within the less perfused area. In 24 patients (22.8%), the HSI transection line had an irregular course and crossed the clinically established resection line. In four cases, HSI disclosed a clinically undetected lesion of the marginal artery. CONCLUSIONS: Intraoperative HSI is safe, well reproducible, and does not disrupt the surgical workflow. It also quantifies bowel surface perfusion. HSI might become an intraoperative guidance tool, potentially preventing postoperative complications.

Tumor cell identification and classification in esophageal adenocarcinoma specimens by hyperspectral imaging.

Esophageal cancer is the sixth leading cause of cancer-related death worldwide. Histopathological confirmation is a key step in tumor diagnosis. Therefore, simplification in decision-making by discrimination between malignant and non-malignant cells of histological specimens can be provided by combination of new imaging technology and artificial intelligence (AI). In this work, hyperspectral imaging (HSI) data from 95 patients were used to classify three different histopathological features (squamous epithelium cells, esophageal adenocarcinoma (EAC) cells, and tumor stroma cells), based on a multi-layer perceptron with two hidden layers. We achieved an accuracy of 78% for EAC and stroma cells, and 80% for squamous epithelium. HSI combined with machine learning algorithms is a promising and innovative technique, which allows image acquisition beyond Red-Green-Blue (RGB) images. Further method validation and standardization will be necessary, before automated tumor cell identification algorithms can be used in daily clinical practice.

New Intraoperative Imaging Tools and Image-Guided Surgery in Gastric Cancer Surgery.

Innovations and new advancements in intraoperative real-time imaging have gained significant importance in the field of gastric cancer surgery in the recent past. Currently, the most promising procedures include indocyanine green fluorescence imaging (ICG-FI) and hyperspectral imaging or multispectral imaging (HSI, MSI). ICG-FI is utilized in a broad range of clinical applications, e.g., assessment of perfusion or lymphatic drainage, and additional implementations are currently investigated. HSI is still in the experimental phase and its value and clinical relevance require further evaluation, but initial studies have shown a successful application in perfusion assessment, and prospects concerning non-invasive tissue and tumor classification are promising. The application of machine learning and artificial intelligence technologies might enable an automatic evaluation of the acquired image data in the future. Both methods facilitate the accurate visualization of tissue characteristics that are initially indistinguishable for the human eye. By aiding surgeons in optimizing the surgical procedure, image-guided surgery can contribute to the oncologic safety and reduction of complications in gastric cancer surgery and recent advances hold promise for the application of HSI in intraoperative tissue diagnostics.

Using physiological parameters measured by hyperspectral imaging to detect colorectal cancer.

The accurate detection of malignant tissue during colorectal surgery impacts operation outcome. The non-invasive spectral imaging combined with machine learning (ML) methods showed to be promising for tumor identification. However, large spectral range implies large computing time. To reduce the number of features, ML methods (e.g. logistic regression and convolutional neuronal network CNN) were evaluated based on four physiological tissue parameters to automatically classify cancer and healthy mucosa in resected colon tissue. A ROC AUC of 0.81 was achieved with the CNN. This study shows that the use of only specific wavelengths bands can detect cancer.Clinical Relevance- These outcomes support the possibility to automatically classify colon tumor based on physiological parameters calculated using only specific wavelength bands. Hence, future image-guided colorectal surgeries can be performed with real-time multispectral imaging.

Hyperspectral Imaging: A New Intraoperative Tool for Pouch Assessment in Patients Undergoing Restorative Proctocolectomy.

INTRODUCTION: Restorative proctocolectomy with ileal pouch-anal anastomosis (IPAA) is a challenging operation. Especially the mobilization of the pouch into the pelvis can be complex. Adequate perfusion of the pouch is required for optimal healing and functioning. METHODS: With hyperspectral imaging (HSI) wavelengths between 500 and 1,000 nm can be analyzed in addition to visible light and by reflecting patterns. This intraoperative procedure is non-invasive, contact-free, and no contrast medium is needed. Fifteen patients undergoing IPAA were examined prospectively, and the pouch was evaluated by HSI intraoperatively. HSI was measured in standardized fashion at 4 defined locations of the J-pouch. Each measurement took about 10 s. The clinical postoperative course was assessed in all patients and correlated to the intraoperative HSI findings. RESULTS: Mean near-infrared perfusion and oxygenation of patients showed values ≥74% for all defined pouch areas, revealing good blood supply. Three minor anastomotic leaks were detected by standard pouchoscopy in the postoperative course, which could be treated conservatively with endosponge therapy. CONCLUSION: HSI values of perfusion and oxygenation of the IPAA were high. The leak rate is associated with redo procedures. This is reflected by the current literature and most likely related to the higher complexity of the revisional pouch operation. HSI has proved itself as a quick and effective new intraoperative tool to evaluate pouch perfusion objectively and quantitatively.

Automatic Recognition of Colon and Esophagogastric Cancer with Machine Learning and Hyperspectral Imaging.

There are approximately 1.8 million diagnoses of colorectal cancer, 1 million diagnoses of stomach cancer, and 0.6 million diagnoses of esophageal cancer each year globally. An automatic computer-assisted diagnostic (CAD) tool to rapidly detect colorectal and esophagogastric cancer tissue in optical images would be hugely valuable to a surgeon during an intervention. Based on a colon dataset with 12 patients and an esophagogastric dataset of 10 patients, several state-of-the-art machine learning methods have been trained to detect cancer tissue using hyperspectral imaging (HSI), including Support Vector Machines (SVM) with radial basis function kernels, Multi-Layer Perceptrons (MLP) and 3D Convolutional Neural Networks (3DCNN). A leave-one-patient-out cross-validation (LOPOCV) with and without combining these sets was performed. The ROC-AUC score of the 3DCNN was slightly higher than the MLP and SVM with a difference of 0.04 AUC. The best performance was achieved with the 3DCNN for colon cancer and esophagogastric cancer detection with a high ROC-AUC of 0.93. The 3DCNN also achieved the best DICE scores of 0.49 and 0.41 on the colon and esophagogastric datasets, respectively. These scores were significantly improved using a patient-specific decision threshold to 0.58 and 0.51, respectively. This indicates that, in practical use, an HSI-based CAD system using an interactive decision threshold is likely to be valuable. Experiments were also performed to measure the benefits of combining the colorectal and esophagogastric datasets (22 patients), and this yielded significantly better results with the MLP and SVM models.