Laparoscopic cholecystectomy in a swine model using a novel near-infrared fluorescent IV dye (BL-760).

BACKGROUND AND OBJECTIVES: Bile duct injury during laparoscopic cholecystectomy has an incidence rate of 1%-2% and commonly appears under conditions of severe inflammation, adhesion, or unexpected anatomical variations. Despite the difficulties and rising concerns of identifying bile duct during surgeries, surgeons do not have a specific modality to identify bile duct except intraoperative cholangiography. While no biliary-specific fluorescent dye exists for clinical use, our team has previously described the development of a preclinical biliary-specific dye, BL-760. Here, we present our study of laparoscopic cholecystectomy using the fluorescent dye in a swine model. STUDY DESIGN/MATERIALS AND METHODS: With an approval from Institutional Animal Care and Use Committee, two 20-25 kg swine underwent laparoscopic abdominal surgery using a Food and Drug Administration-cleared fluorescent laparoscopic system. Images of the liver and gallbladder were taken both before and after intravenous injection of the novel fluorescent dye. The dye was dosed at 60 µg/kg and injected via the ear vein. The amount of time taken to visualize fluorescence in the biliary tract was measured. Fluorescent signal was observed after injection, and target-to-background ratio (TBR) of the biliary tract to surrounding cystic artery and liver parenchyma was measured. RESULTS: Biliary tract visualization under fluorescent laparoscopy was achieved within 5 min after the dye injection without any adverse effects. Cystic duct and extrahepatic duct were clearly visualized and identified with TBR values of 2.19 and 2.32, respectively, whereas no fluorescent signal was detected in liver. Cystic duct and artery were successfully ligated by an endoscopic clip applier with the visual assistance of highlighted biliary tract images. Laparoscopic cholecystectomy was completed within 30 min in each case without any complications. CONCLUSIONS: BL-760 is a novel preclinical fluorescent dye useful for intraoperative identification and visualization of biliary tract. Such fluorescent dye that is exclusively metabolized by liver and rapidly excreted into biliary tract would be beneficial for all types of hepato-biliary surgeries. With the validation of additional preclinical data, this novel dye has potential to be a valuable tool to prevent any iatrogenic biliary injuries and/or bile leaks during laparoscopic abdominal and liver surgeries.

Explaining the Rationale of Deep Learning Glaucoma Decisions with Adversarial Examples.

PURPOSE: To illustrate what is inside the so-called black box of deep learning models (DLMs) so that clinicians can have greater confidence in the conclusions of artificial intelligence by evaluating adversarial explanation on its ability to explain the rationale of DLM decisions for glaucoma and glaucoma-related findings. Adversarial explanation generates adversarial examples (AEs), or images that have been changed to gain or lose pathologic characteristic-specific traits, to explain the DLM's rationale. DESIGN: Evaluation of explanation methods for DLMs. PARTICIPANTS: Health screening participants (n = 1653) at the Seoul National University Hospital Health Promotion Center, Seoul, Republic of Korea. METHODS: We trained DLMs for referable glaucoma (RG), increased cup-to-disc ratio (ICDR), disc rim narrowing (DRN), and retinal nerve fiber layer defect (RNFLD) using 6430 retinal fundus images. Surveys consisting of explanations using AE and gradient-weighted class activation mapping (GradCAM), a conventional heatmap-based explanation method, were generated for 400 pathologic and healthy patient eyes. For each method, board-trained glaucoma specialists rated location explainability, the ability to pinpoint decision-relevant areas in the image, and rationale explainability, the ability to inform the user on the model's reasoning for the decision based on pathologic features. Scores were compared by paired Wilcoxon signed-rank test. MAIN OUTCOME MEASURES: Area under the receiver operating characteristic curve (AUC), sensitivities, and specificities of DLMs; visualization of clinical pathologic changes of AEs; and survey scores for locational and rationale explainability. RESULTS: The AUCs were 0.90, 0.99, 0.95, and 0.79 and sensitivities were 0.79, 1.00, 0.82, and 0.55 at 0.90 specificity for RG, ICDR, DRN, and RNFLD DLMs, respectively. Generated AEs showed valid clinical feature changes, and survey results for location explainability were 3.94 ± 1.33 and 2.55 ± 1.24 using AEs and GradCAMs, respectively, of a possible maximum score of 5 points. The scores for rationale explainability were 3.97 ± 1.31 and 2.10 ± 1.25 for AEs and GradCAM, respectively. Adversarial example provided significantly better explainability than GradCAM. CONCLUSIONS: Adversarial explanation increased the explainability over GradCAM, a conventional heatmap-based explanation method. Adversarial explanation may help medical professionals understand more clearly the rationale of DLMs when using them for clinical decisions.

A Biliary Tract-Specific Near-Infrared Fluorescent Dye for Image-Guided Hepatobiliary Surgery.

Despite advances, visual inspection, palpation, and intraoperative ultrasound remain the most utilized tools during surgery today. A particularly challenging issue is the identification of the biliary system due to its complex architecture partially embedded within the liver. Fluorescence guided surgical interventions, particularly using near-infrared (NIR) wavelengths, are an emerging approach for the real-time assessment of the hepatobiliary system. However, existing fluorophores, such as the FDA-approved indocyanine green (ICG), have significant limitations for rapid and selective visualization of bile duct anatomy. Here we report a novel NIR fluorophore, BL (Bile Label)-760, which is exclusively metabolized by the liver providing high signal in the biliary system shortly after intravenous administration. This molecule was identified by first screening a small set of known heptamethine cyanines including clinically utilized agents. After finding that none of these were well-suited, we then designed and tested a small series of novel dyes within a prescribed polarity range. We validated the molecule that emerged from these efforts, BL-760, through animal studies using both rodent and swine models employing a clinically applicable imaging system. In contrast to ICG, BL-760 fluorescence revealed a high target-to-background ratio (TBR) of the cystic duct relative to liver parenchyma 5 min after intravenous injection. During hepatic resection surgery, intrahepatic ducts were clearly highlighted, and bile leakage was easily detected. In conclusion, BL-760 has highly promising properties for intraoperative navigation during hepatobiliary surgery.

Sentinel node navigation surgery using near-infrared indocyanine green fluorescence in early gastric cancer.

BACKGROUND: The aim of this study was to evaluate the feasibility of indocyanine green (ICG) fluorescent method for sentinel lymph node detection in early gastric cancer. METHODS: Between December 2012 and December 2014, 28 cases of pilot examination were performed at Seoul National University Bundang Hospital. Advanced version of multispectral fluorescence organoscope was used to identify sentinel node by quantitative estimation of ICG fluorescent signal intensity. Sensitivity, specificity, false positive value were analyzed and compared with dual tracer method. RESULTS: A total of 443 lymph nodes in 28 cases were examined and 184 sentinel nodes (41.5%) were identified by dual tracer method. The sensitivity using near-infrared ICG method was 98.9%. The specificity was 76.0% and false positive rate was 25.4% compared with dual tracer method. The adequate threshold for sentinel node detection was considered as 10% of maximum signal intensity. CONCLUSION: New near-infrared ICG fluorescent method could be a promising protocol for sentinel node navigation surgery in early gastric cancer.

Association of Cardiovascular Mortality and Deep Learning-Funduscopic Atherosclerosis Score derived from Retinal Fundus Images.

PURPOSE: The prediction of atherosclerosis using retinal fundus images and deep learning has not been shown possible. The purpose of this study was to develop a deep learning model which predicted atherosclerosis by using retinal fundus images and to verify its clinical implications by conducting a retrospective cohort analysis. DESIGN: Retrospective cohort study. METHODS: The database at the Health Promotion Center of Seoul National University Hospital (HPC-SNUH) was used. The deep learning model was trained using 15,408 images to predict carotid artery atherosclerosis, which was named the deep-learning funduscopic atherosclerosis score (DL-FAS). A retrospective cohort was constructed of participants 30-80 years old who had completed elective health examinations at HPC-SNUH. Using DL-FAS as the main exposure, participants were followed for the primary outcome of death due to CVD until Dec. 31, 2017. RESULTS: For predicting carotid artery atherosclerosis among subjects, the model achieved an area under receiver operating curve (AUROC) and area under the precision-recall curve (AUPRC), accuracy, sensitivity, specificity, positive and negative predictive values of 0.713, 0.569, 0.583, 0.891, 0.404, 0.465, and 0.865 respectively. The cohort consisted of 32,227 participants, 78 cardiovascular disease (CVD) deaths, and 7.6-year median follow-up visits. Those with DL-FAS greater than 0.66 had an increased risk of CVD deaths compared to those with DL-FAS <0.33 (hazard ratio: 8.33; 95% confidence interval [CI], 3.16-24.7). Risk association was significant among intermediate and high Framingham risk score (FRS) subgroups. The DL-FAS improved the concordance by 0.0266 (95% CI, 0.0043-0.0489) over the FRS-only model. The relative integrated discrimination index was 20.45% and net reclassification index was 29.5%. CONCLUSIONS: A deep learning model was developed which could predict atherosclerosis from retinal fundus images. The resulting DL-FAS was an independent predictor of CVD deaths when adjusted for FRS and added predictive value over FRS.

Dual-Channel Endoscopic Indocyanine Green Fluorescence Angiography for Clipping of Cerebral Aneurysms.

OBJECTIVE: Neuroendoscopy is useful for assessing status of perforators, parent arteries, and aneurysms beyond the straight line of microscopic view during aneurysm clipping. We aimed to evaluate the clinical usefulness of our endoscopic indocyanine green angiography (ICGA) system, which can simultaneously display visible light and indocyanine green fluorescent images. METHODS: Surgical clipping of 16 unruptured aneurysms in 10 patients was performed via the keyhole approach. Using our endoscopic ICGA and commercial microscopic ICGA systems, we prospectively compared 10 targeted cerebral aneurysms at the posterior communicating (n = 4) and anterior choroidal (n = 6) arteries. RESULTS: Microscopic ICGA and endoscopic ICGA were feasible during surgery. Microscopic ICGA displayed 50% of branch orifices, 100% of branch trunks, and 20% of exact clip positions, whereas endoscopic ICGA showed 100% of these. Based on endoscopic ICGA findings such as incomplete clipping and compromise of parent arteries or branches, clips were repositioned in 2 cases, and additional clips were applied in 2 cases. Complete occlusion and residual neck states were achieved in 6 and 4 aneurysms after surgery. There were no neurologic deficits within 3 months after surgery except for frontalis palsy and anosmia in each patient. CONCLUSIONS: The endoscopic ICGA system with dual imaging of visible light and indocyanine green fluorescence was very useful for assessing geometry of aneurysms and surrounding vessels before clipping and for evaluating completeness of clip position after clipping.

Preliminary evaluation of the use of a CDMA-based emergency telemedicine system.

We have conducted a preliminary evaluation of the use of a code division multiple access (CDMA)-based emergency telemedicine system to be used by emergency rescuers providing first-aid treatment for patients. The prototype system included five instrumentation modules for measuring non-invasive arterial blood pressure (NIBP), arterial oxygen saturation (SpO(2)), six-channel electrocardiogram (ECG), blood glucose concentration and body temperature. The patient data were transferred to a doctor's PC through CDMA and TCP/IP networks using an embedded personal digital assistant (PDA) phone. Performance tests in the laboratory showed that the system provided reliable values with error ranges within +/-1% for heart rate, +/-5 mmHg for NIBP, +/-2% for SpO(2) and +/-1% for glucose. The feasibility of the prototype system was then evaluated with 15 real emergency patients on Jeju Island over a two-month period. Measured data were transmitted from a moving ambulance to the emergency medical centre without significant CDMA connection loss or transmission errors. The average transfer time was 8 min. Four emergency doctors and 11 rescuers completed a questionnaire. There were favourable reviews from the users.

A novel imaging platform for non-invasive screening of abnormal glucose tolerance.

Optical measurement of skin auto-fluorescence (SAF), most likely emanating from accumulated advanced glycation end-products (AGEs), has been proposed for the noninvasive diagnosis of glucose intolerance in clinical settings. Here, we developed a novel imaging system with transmission geometry for SAF measurement and compared its diagnostic performance in a Korean population.

Cation-selective electropreconcentration.

A cation-selective microfluidic sample preconcentration system is described. The cation sample was electropreconcentrated using a reversed-direction electroosmotic flow (EOF) and an anion-permselective filter, where an electric double layer (EDL) overlap condition existed. The anion-permselective filter between microchannels was fabricated by three different methods: 1) extending a positively charged, nanoporous, polymer membrane by photopolymerization of poly(diallyldimethylammonium chloride) (PDADMAC); 2) etching a nanochannel and then coating it with a positively-charged monomer, N-[3-(trimethoxysilyl)propyl]-N'-(4-vinylbenzyl)ethylenediamine hydrochloride (TMSVE); and, 3) etching a nanochannel and then coating it with a positively-charged, pre-formed polymer, polyE-323. The EOF direction in the microchannel was reversed by both TMSVE and polyE-323 coatings. The cation-selective preconcentration was investigated using charged fluorescent dyes and tetramethylrhodamine isothiocyanate (TRITC)-tagged peptides/proteins. The preconcentration in the three different systems was compared with respect to efficiency, dependence on buffer concentration and pH, tolerable flow rate, and sample adsorption. Both TMSVE- and polyE-323-coated nanochannels showed robust preconcentration at high flow rates, whereas the PDADMAC membrane maintained anion-permselectivity at higher buffer concentrations. The TMSVE-coated nanochannels showed a more stable preconcentration process, whereas the polyE-323-coated nanochannels showed a lower peptide sample adsorption and robust efficiency under a wide range of buffer pHs. The system described here can potentially be used for the preconcentration of cationic peptides/proteins on microfluidic devices for subsequent analyses.

Automatic stress-relieving music recommendation system based on photoplethysmography-derived heart rate variability analysis.

This paper presents an automatic stress-relieving music recommendation system (ASMRS) for individual music listeners. The ASMRS uses a portable, wireless photoplethysmography module with a finger-type sensor, and a program that translates heartbeat signals from the sensor to the stress index. The sympathovagal balance index (SVI) was calculated from heart rate variability to assess the user's stress levels while listening to music. Twenty-two healthy volunteers participated in the experiment. The results have shown that the participants' SVI values are highly correlated with their prespecified music preferences. The sensitivity and specificity of the favorable music classification also improved as the number of music repetitions increased to 20 times. Based on the SVI values, the system automatically recommends favorable music lists to relieve stress for individuals.

Implementation of a real-time multi-channel gateway server in ubiquitous integrated biotelemetry system for emergency care (UIBSEC).

We developed a gateway server to support various types of bio-signal monitoring devices for ubiquitous emergency healthcare in a reliable, effective, and scalable way. The server provides multiple channels supporting real-time N-to-N client connections. We applied our system to four types of health monitoring devices including a 12-channel electrocardiograph (ECG), oxygen saturation (SpO(2)), and medical imaging devices (a ultrasonograph and a digital skin microscope). Different types of telecommunication networks were tested: WIBRO, CDMA, wireless LAN, and wired internet. We measured the performance of our system in terms of the transmission rate and the number of simultaneous connections. The results show that the proposed network communication strategy can be successfully applied to the ubiquitous emergency healthcare service by providing a fast rate enough for real-time video transmission and multiple connections among patients and medical personnel.

Real-time medical control using a wireless audio-video transmission device in a pre-hospital emergency service in Korea.

We developed a hands-free portable device which can provide two-way, real-time audio and video communication between hospital doctors and emergency medical technicians (EMTs) providing pre-hospital care. The device was based on an ultra mobile PC with a camera, a microphone/earphone set and a WIBRO modem for wireless Internet connection at an average data transmission rate of 1 Mbit/s. Feasibility tests were conducted in 55 real emergency situations over a period of three months at five different Rescue Centres in Seoul. Successful communication between an EMT and a doctor was achieved in 46 cases (84%). The device showed acceptable performance in terms of audio/video transmission time delays and maximum transmitted video frame rates, both outdoors, inside a building and in a moving vehicle at 70 km/h. Eight control centre staff and 11 EMTs who used the device completed a questionnaire. Despite acceptable basic performance, the device was found to be limited in terms of the contribution it made to the medical control of EMTs. However, improvements in device performance should produce higher quality pre-hospital emergency medical care in the future.

Ubiquitous monitoring system for chronic obstructive pulmonary disease and heart disease patients.

The Purpose of this paper is to report a code division multiple access (CDMA) based wireless device that is able to measure pulse oxygen saturation(SpO2) and Electrocardiogram(ECG) during chronic obstructive pulmonary disease and heart disease patients' daily life at home. Global Positioning System (GPS) and a specific database system are also included in this device to trace patient's location. Unlike the pre-reported devices using wireless network like Bluetooth or Zigbee which has a limited connection area, the developed system enables the literally ubiquitous service in reporting the patient's data to their doctors at any time and any place using the CDMA-based cellular phone network.

Implementation of the relay server for the ubiquitous integrated biotelemetry system for emergency care (UIBSEC) based on 3.5G HSDPA technology.

When an emergent case occurs, to give proper and immediate emergency treatment to a patient is as important as to transfer him or her to the hospital as early as possible. For this circumstance where first aid usually happens in an ambulance, we developed the Ubiquitous Integrated Biotelemetry System for Emergency Care (UIBSEC) based on wireless 3.5G HSDPA module in order to provide more proper and active medical care to a patient. For this system we performed experimental tests on the HSDPA module to measure practical performance values with respect to throughput and RTT (round-trip-time) and compared the result with the theoretical specifications. In the course of developing this system we created the C#-based Network Solution Library (NSL) which helps drastically reducing the overhead. We applied this library to components of the UIBSEC - two kinds of clients and one central server - and confirmed that it works smoothly in wireless networks and it is useful in the future development process of network applications.

Preliminary study of motion artifact rejection for NIBP measurement in an ambulance.

Motion artifact resulting from patient's movement is a significant source for disturbing accurate noninvasive blood pressure (NIBP) measurement. In an ambulance, patients are exposed to unstable circumstances due to vehicle's movement and vibration during emergency transportation. Since NIBP is indirectly measured using oscillometry based on the change of cuff pressure, it can be affected by motion artifact much more than other biosignals. In this paper, we developed a new NIBP system with improved accuracy by measuring acceleration of the system caused by patient's motion. The NIBP module including a 3-axis accelerometer was directly mounted on a cuff to minimize the interference induced through connecting tube. The results show that the proposed NIBP system has the capability to reject the interference of motion artifact effectively in an ambulance.

HSDPA (3.5G)-based ubiquitous integrated biotelemetry system for emergency care.

We have developed the second prototype system of Ubiquitous Integrated Biotelemetry System for Emergency Care(UIBSEC) using a HSDPA(High Speed Downlink Packet Access) modem to be used by emergency rescuers to get directions from medical doctors in providing emergency medical services for patients in ambulance. Five vital bio-signal instrumentation modules have been implemented, which include noninvasive arterial blood pressure (NIBP), arterial oxygen saturation (SaO2), 6-channel electro-cardiogram(ECG), blood glucose level, and body temperature and real-time motion picture of the patient and GPS information are also taken. Measured patient data, captured motion picture and GPS information are then transferred to a doctor's PC through the HSDPA and TCP/IP networks using stand-alone HSDPA modem. Most prominent feature of the developed system is that it is based on the HSDPA backbone networks available in Korea now, through which we will be able to establish a ubiquitous emergency healthcare service system.