Baseline Radiomic Signature to Estimate Overall Survival in Patients With NSCLC.

INTRODUCTION: We aimed to define a baseline radiomic signature associated with overall survival (OS) using baseline computed tomography (CT) images obtained from patients with NSCLC treated with nivolumab or chemotherapy. METHODS: The radiomic signature was developed in patients with NSCLC treated with nivolumab in CheckMate-017, -026, and -063. Nivolumab-treated patients were pooled and randomized to training, calibration, or validation sets using a 2:1:1 ratio. From baseline CT images, volume of tumor lesions was semiautomatically segmented, and 38 radiomic variables depicting tumor phenotype were extracted. Association between the radiomic signature and OS was assessed in the nivolumab-treated (validation set) and chemotherapy-treated (test set) patients in these studies. RESULTS: A baseline radiomic signature was identified using CT images obtained from 758 patients. The radiomic signature used a combination of imaging variables (spatial correlation, tumor volume in the liver, and tumor volume in the mediastinal lymph nodes) to output a continuous value, ranging from 0 to 1 (from most to least favorable estimated OS). Given a threshold of 0.55, the sensitivity and specificity of the radiomic signature for predicting 3-month OS were 86% and 77.8%, respectively. The signature was identified in the training set of patients treated with nivolumab and was significantly associated (p < 0.0001) with OS in patients treated with nivolumab or chemotherapy. CONCLUSIONS: The radiomic signature provides an early readout of the anticipated OS in patients with NSCLC treated with nivolumab or chemotherapy. This could provide important prognostic information and may support risk stratification in clinical trials.

Identification of Non-Small Cell Lung Cancer Sensitive to Systemic Cancer Therapies Using Radiomics.

PURPOSE: Using standard-of-care CT images obtained from patients with a diagnosis of non-small cell lung cancer (NSCLC), we defined radiomics signatures predicting the sensitivity of tumors to nivolumab, docetaxel, and gefitinib. EXPERIMENTAL DESIGN: Data were collected prospectively and analyzed retrospectively across multicenter clinical trials [nivolumab, n = 92, CheckMate017 (NCT01642004), CheckMate063 (NCT01721759); docetaxel, n = 50, CheckMate017; gefitinib, n = 46, (NCT00588445)]. Patients were randomized to training or validation cohorts using either a 4:1 ratio (nivolumab: 72T:20V) or a 2:1 ratio (docetaxel: 32T:18V; gefitinib: 31T:15V) to ensure an adequate sample size in the validation set. Radiomics signatures were derived from quantitative analysis of early tumor changes from baseline to first on-treatment assessment. For each patient, 1,160 radiomics features were extracted from the largest measurable lung lesion. Tumors were classified as treatment sensitive or insensitive; reference standard was median progression-free survival (NCT01642004, NCT01721759) or surgery (NCT00588445). Machine learning was implemented to select up to four features to develop a radiomics signature in the training datasets and applied to each patient in the validation datasets to classify treatment sensitivity. RESULTS: The radiomics signatures predicted treatment sensitivity in the validation dataset of each study group with AUC (95 confidence interval): nivolumab, 0.77 (0.55-1.00); docetaxel, 0.67 (0.37-0.96); and gefitinib, 0.82 (0.53-0.97). Using serial radiographic measurements, the magnitude of exponential increase in signature features deciphering tumor volume, invasion of tumor boundaries, or tumor spatial heterogeneity was associated with shorter overall survival. CONCLUSIONS: Radiomics signatures predicted tumor sensitivity to treatment in patients with NSCLC, offering an approach that could enhance clinical decision-making to continue systemic therapies and forecast overall survival.

Optoacoustic imaging of the prostate: development toward image-guided biopsy.

Optoacoustic (OA) tomography has demonstrated utility in identifying blood-rich malignancies in breast tissue. We describe the development and characterization of a laser OA imaging system for the prostate (LOIS-P). The system consists of a fiber-coupled Q-switched laser operating at 757 nm, a commercial 128-channel ultrasonic probe, a digital signal processor, and software that uses the filtered radial back-projection algorithm for image reconstruction. The system is used to reconstruct OA images of a blood-rich lesion induced in vivo in a canine prostate. OA images obtained in vivo are compared to images acquired using ultrasound, the current gold standard for guiding biopsy of the prostate. Although key structural features such as the urethra could be identified with both imaging techniques, a bloody lesion representing a highly vascularized tumor could only be clearly identified in OA images. The advantages and limitations of both forward and backward illumination modes are also evaluated by collecting OA images of phantoms simulating blood vessels within tissue. System resolution is estimated to be 0.2 mm in the radial direction of the acoustic array. The minimum detectable pressure signal is 1.83 Pa. Our results encourage further development toward a dual-modality OA/ultrasonic system for prostate imaging and image-guided biopsy.

Vibrating interventional device detection using real-time 3-D color Doppler.

Ultrasound image guidance of interventional devices during minimally invasive surgery provides the clinician with improved soft tissue contrast while reducing ionizing radiation exposure. One problem with ultrasound image guidance is poor visualization of the device tip during the clinical procedure. We have described previously guidance of several interventional devices using a real-time 3-D (RT3-D) ultrasound system with 3-D color Doppler combined with the ColorMark technology. We then developed an analytical model for a vibrating needle to maximize the tip vibrations and improve the reliability and sensitivity of our technique. In this paper, we use the analytical model and improved radiofrequency (RF) and color Doppler filters to detect two different vibrating devices in water tank experiments as well as in an in vivo canine experiment. We performed water tank experiments with four different 3- D transducers: a 5 MHz transesophageal (TEE) probe, a 5 MHz transthoracic (TTE) probe, a 5 MHz intracardiac catheter (ICE) transducer, and a 2.5 MHz commercial TTE probe. Each transducer was used to scan an aortic graft suspended in the water tank. An atrial septal puncture needle and an endomyocardial biopsy forceps, each vibrating at 1.3 kHz, were inserted into the vascular graft and were tracked using 3-D color Doppler. Improved RF and wall filters increased the detected color Doppler sensitivity by 14 dB. In three simultaneous planes from the in vivo 3-D scan, we identified both the septal puncture needle and the biopsy forceps within the right atrium using the 2.5 MHz probe. A new display filter was used to suppress the unwanted flash artifact associated with physiological motion.

Analysis of a vibrating interventional device to improve 3-D colormark tracking.

Ultrasound guidance of interventional devices during minimally invasive surgical procedures has been investigated by many researchers. Previously, we extended the methods used by the Colormark tracking system to several interventional devices using a real-time, three-dimensional (3-D) ultrasound system. These results showed that we needed to improve the efficiency and reliability of the tracking. In this paper, we describe an analytical model to predict the transverse vibrations along the length of an atrial septal puncture needle to enable design improvements of the tracking system. We assume the needle can be modeled as a hollow bar with a circular cross section with a fixed proximal end and a free distal end that is suspended vertically to ignore gravity effects. The initial results show an ability to predict the natural nodes and antinodes along the needle using the characteristic equation for free vibrations. Simulations show that applying a forcing function to the device at a natural antinode yields an order of magnitude larger vibration than when driving the device at a node. Pulsed wave spectral Doppler data was acquired along the distal portion of the needle in a water tank using a 2-D matrix array transesophageal echocardiography probe. This data was compared to simulations of forced vibrations from the model. These initial results suggest that the model is a good first order approximation of the vibrating device in a water tank. It is our belief that knowing the location of the natural nodes and antinodes will improve our ability to drive the device to ensure the vibrations at the proximal end will reach the tip of the device, which in turn should improve our ability to track the device in vivo.

Autonomous surgical robotics using 3-D ultrasound guidance: feasibility study.

The goal of this study was to test the feasibility of using a real-time 3D (RT3D) ultrasound scanner with a transthoracic matrix array transducer probe to guide an autonomous surgical robot. Employing a fiducial alignment mark on the transducer to orient the robot's frame of reference and using simple thresholding algorithms to segment the 3D images, we tested the accuracy of using the scanner to automatically direct a robot arm that touched two needle tips together within a water tank. RMS measurement error was 3.8% or 1.58 mm for an average path length of 41 mm. Using these same techniques, the autonomous robot also performed simulated needle biopsies of a cyst-like lesion in a tissue phantom. This feasibility study shows the potential for 3D ultrasound guidance of an autonomous surgical robot for simple interventional tasks, including lesion biopsy and foreign body removal.

3-D ultrasound guidance of surgical robotics: a feasibility study.

Laparoscopic ultrasound has seen increased use as a surgical aide in general, gynecological, and urological procedures. The application of real-time, three-dimensional (RT3D) ultrasound to these laparoscopic procedures may increase information available to the surgeon and serve as an additional intraoperative guidance tool. The integration of RT3D with recent advances in robotic surgery also can increase automation and ease of use. In this study, a 1-cm diameter probe for RT3D has been used laparoscopically for in vivo imaging of a canine. The probe, which operates at 5 MHz, was used to image the spleen, liver, and gall bladder as well as to guide surgical instruments. Furthermore, the three-dimensional (3-D) measurement system of the volumetric scanner used with this probe was tested as a guidance mechanism for a robotic linear motion system in order to simulate the feasibility of RT3D/robotic surgery integration. Using images acquired with the 3-D laparoscopic ultrasound device, coordinates were acquired by the scanner and used to direct a robotically controlled needle toward desired in vitro targets as well as targets in a post-mortem canine. The rms error for these measurements was 1.34 mm using optical alignment and 0.76 mm using ultrasound alignment.

Real-time 3D color flow Doppler for guidance of vibrating interventional devices.

The goal of this investigation was to examine the feasibility of guiding interventional devices using piezoelectric buzzers to create velocity sources, which were imaged and tracked with real-time 3D color flow Doppler. The interventional devices examined in this study included a pacemaker lead, Brockenbrough needle for cardiac septal puncture, cardiac guidewire and radiofrequency ablation needles for cancer therapy. Each was mechanically coupled to a piezoelectric buzzer and was imaged using a commercial real-time 3D ultrasound system with either a 2.5 MHz matrix array transducer or a 5 MHz, 22 F catheter transducer equipped with a tool port. In vitro images acquired in tissue phantoms, excised liver with a 'tumor' target and an excised sheep heart show strong vibration signals in 3D color flow Doppler, enabling real-time tracking and guidance of all the devices in three dimensions. In a sheep model, in vivo tracking of the pacing lead was performed in the superior vena cava as well as the right atrium using RT3D color flow Doppler images. The vibrating rf ablation needles were guided through the liver toward "tumor" targets in vivo with real-time 3D color flow Doppler images.