Hi all researchers,

As known in most of morphological studies, the facial soft-tissue thicknesses are interested to apply in reconstruction, segmentation, and partly in ALADA-based CT exposure optimization in forensic sciences, anthropology, oral-facial medicines, and CT physics. In the practical studies, factors of anthropological metrics, genders, ages/ maturity, body-mass-indices/ nutrition conditions, socio-geographical influences, etc are accounted and statistically analyzed. Factually, that is not a completeness of FSTS for the relevant study about head-and-neck, oral and dentition anatomy (HaNODA), due to study objectives, proposed measurement methods, and limitation of study populations.

Those data is not fully applicable in soft-tissue detection and classification application in medical diagnostics, treatment planning, forensic reconstruction, anthropological morphology reconstruction, and also for diagnostic imaging devices design. So I would like to have your advices in this field what should be the best proposal for measurement method and incorporation of the existed databases.

There are several existed databases and studies, that I am interested in and using for my projects, listed as followed:

1. De Greef, S. et al (2006) proposed a large-scale study on Caucasian population (976 patients, classified by genders, ages (18 - 65 yo) and BMIs), using 31 FSTS landmarks (10 monotonical distributed on facial midline, and 21 bilaterally distributed), which is measured by B-mode US scanner. Reference link: Article Large-scale in-vivo Caucasian soft tissue thickness database...

2. Hwang, H.-S. et al (2012) proposed a CBCT study on Korean adults (a pilot-study population of 100 patients, ratio of males-to-females is 1:1), based on the anatomical landmarking proposal of De Greef, S. et al (2006). Reference link: Article Facial Soft Tissue Thickness Database for Craniofacial Recon...

3. Weinberg, S.M. et al (FaceBase.org) (2009 - 2019) 3D facial norm database of European caucasians, for craniofacial researchers, using both in-vivo conventional measurement and in-simuli measurement on 3D white light structured photometric scanners. The 3D scanned (point-clouded) meshes are use for metrics of morphic contouring of head and neck anatomy (conventional calipering (in-vivo (5) and in-simuli (9)), 3D measurement based on 3D landmarks (8 monotonical distributed on facial midline, and 8 bilaterally distributed )), incorporated with weights and heights. Reference link: The FaceBase Consortium: a comprehensive resource for craniofacial researchers. Brinkley JF, Fisher S, Harris MP, Holmes G, Hooper JE, Jabs EW, Jones KL, Kesselman C, Klein OD, Maas RL, Marazita ML, Selleri L, Spritz RA, van Bakel H, Visel A, Williams TJ, Wysocka J; FaceBase Consortium, Chai Y. Development. 2016 Jul 15;143(14):2677-88. doi: 10.1242/dev.135434. Epub 2016 Jun 10.

https://www.facebase.org/facial_norms/

According to the reviews about the field of Stephan C.N et al (2015 and 2019) and Olate, S. et al (2017), I think that the modern approach for this case might be the fusion of CBCT database, 3D optic structured light scanner databases, and 3D morphological nominal meshing. Unfortunately, there has no studies in this approach yet. I would like to find the relevant database and using the conventional database to estimate the new linear regression equations for problem, as my proposal:

t(T_i) = (b_dimophism) * (t_muy) + (b_aging) * (t_age) + (b_BMI) * (t_BMI) + (lambda_TEMcorr) * (E_rTEM)

in which:

1. t(T_i) is the thickness of the anatomy structure along the beam transmission line from source to detector (in mm).

2. T_i is a compositional of the transmission media {T} of the beam travelling, composed by anatomical sub-media, s.t.: {T} = union of {T_i}. I assumed as i = 4, as followed thresholded:

T_1: ambience

T_2: soft-tissue structure

T_3: hard-tissue structure

T_4: supra-high artifact

3. b_dimorphism is a coeff of dimorphic factors, influenced to the median thickness t_muy.

4. b_aging is a coeff of aging, influences to the median thickness of t_age.

5. b_BMI is a coeff of nutrition conditions, influences to the median thickness of t_BMI.

6. lambda_TEMcorr is correction factor for correcting the technical error metrics

7. E_rTEM is the average error of the relative technical error metrics due to differences of measurement methods.

I already did the computation of the possible attenuation coeff per each peripheral projective features of Caucasian ODA (cODA), based on De Greef study and ICRU Report 44. It is partially use for my proposed selective anatomical Analysis reconstruction algorithm (based on featured quadrants of full-scan-revolution) (published), and will be used for my second project on cross-junction of FSTS segmentation (in-progressing). Even the data boundaries But I still question how to efficiently eliminate the "overlapping" and the "realiability sparsity" of the b_dimorphism in terms of ages, anatomical regional differentiations (of mid-facial and lower facial (lips and chin) regions) and rTEM.

Looking forward to your advices.

Regards,

Allan.