Most pain shapes were drawn on a TMT 2D pain drawing program (with front and back views of the human body) and the calculated pain centroids (the mathematical center of each pain shape) were then mapped to TMT's 3D human body model. Data were reviewed to confirm comparable representation on the left and right sides of the body; data were then bilateralized with left-sided centroids being added to the comparable right side of the body, and vice versa.

The "Chronic Pain" link combines data from people with multiple types of pain (lower back/sciatica, knee, shoulder, hip, distal peripheral neuropathy, wrist pain/carpal tunnel syndrome, headache, jaw pain /TMJ, toothache). The Peripheral Neuropathy link includes data primarily from people meeting the criteria of Portenoy et al (2005) for neuropathic pain and were then culled to represent only peripheral limb pain (forearm/wrist/hand and lower leg/ankle/foot), probably representing small fiber peripheral neuropathy. The Bone Metastasis Pain data (courtesy of Charles Cleeland of the MD Anderson Cancer Center, U. Texas, Houston) is derived from pain drawings made using the Brief Pain Inventory (BPI) and then transferred to the TMT 2D coordinate space and then to the TMT 3D coordinate space.

The "Bone Metastasis Location on Bone Scan" data was obtained by mapping bone scans with bone metastases (primarily from US academic centers) to the common coordinate space of the TMT 2D drawing model (Note that the 3D representation of the 2D bone scan metastasis data is provided purely for comparison with the front and back views of the pain data in bone metastasis patients, since the 3D model assumes that the front/back view data is on the body surface, which assumption is not met by the internally located bony skeleton).

Pending full analysis of the data, the following conclusions seems reasonable:
o There are six centers of knee pain centroids (lateral, central and medial pain in both front and back views) as shown by kernel density estimation and 2D point cluster segmentation using mixture modeling and expectation maximization.
o Sciatica pain occurs primarily in the S1 distribution (whereas the literature usually just says that L5 + S1 accounts for 95% of sciatica).
o Shoulder pain is concentrated near the shoulder joint capsule and the scapula.
o Hip pain is concentrated antero-lateral to the hip joint.
o For distal peripheral neuropathy pain (subjects primarily meeting Portenoy diagnostic neuropathy criteria or with a previous diagnosis of peripheral neuropathy from their healthcare provider, and in whom pain centroids were present in the distal limbs), hand/wrist/forearm pain is primarily on the medial side whereas lower limb pain centroids are centrally distributed in the front and back of the feet/ankles/lower leg (normally representing involvement of the entire anterior, posterior, lateral and medial areas at a given vertical level).
o Wrist pain in subjects most of whom met multiple diagnostic criteria for carpal tunnel syndrome, is primarily in the anterior and posterior sensory distribution of the median nerve.
o Headache is primarily located on the forehead and the back of the head just above the neck.
o Jaw pain in subjects most of whom met multiple diagnostic criteria for temporo-mandibular joint syndrome (TMJ) is primarily located over the temporo-mandibular joint.
o Toothache is primarily located over the upper and lower molars.

TMT’s triangulation-based mapping technology provides a powerful and flexible means for integrated analysis of disparate databases of biological shapes. Spatial distribution patterns of different variables can be compared within the same coordinate space. Thus, for bone metastasis data we compared the location of bony metastases on individual patient bone scans, pain in metastasis patients, pain in non-metastasis patients, body surface outlines, and bony skeleton outlines – all in the same coordinate space.
• The TMT bone scan analysis provides a unique quantitative display of the distribution of bone metastases. Each bone scan (obtained primarily from US academic centers) was individually mapped to the TMT coordinate space (a rather laborious procedure). The TMT bone scan map appears consistent with a literature review of articles providing textual descriptions of the spatial distribution of bone metastases.
• The MDA pain drawing database in bone mets patients appears consistent with the aggregate distribution of bony metastases.
• A comparison of the MDA pain drawing database in bone mets patients appears to show (versus pain in non-cancer subjects) 1) a higher prevalence of pain in the middle/upper spine, the rib cage, and the femur and tibia, and 2) a lower prevalence of pain in the forearm/wrist/hand and the knee. Pain in the lower back, shoulder area and head are common in both patients with bone metastases and subjects with non-metastasis pain.